date
timestamp[ns]date 2023-05-05 00:00:00
2025-03-25 00:00:00
| arxiv_id
stringlengths 10
10
| title
stringlengths 8
177
| authors
sequencelengths 1
942
| github
stringlengths 0
116
| abstract
stringlengths 165
1.92k
|
|---|---|---|---|---|---|
2025-03-24T00:00:00 | 2503.16905 | MAPS: A Multi-Agent Framework Based on Big Seven Personality and Socratic Guidance for Multimodal Scientific Problem Solving | [
"Jian Zhang",
"Zhiyuan Wang",
"Zhangqi Wang",
"Xinyu Zhang",
"Fangzhi Xu",
"Qika Lin",
"Rui Mao",
"Erik Cambria",
"Jun Liu"
] | Multimodal scientific problems (MSPs) involve complex issues that require the integration of multiple modalities, such as text and diagrams, presenting a significant challenge in artificial intelligence. While progress has been made in addressing traditional scientific problems, MSPs still face two primary issues: the challenge of multi-modal comprehensive reasoning in scientific problem-solving and the lack of reflective and rethinking capabilities. To address these issues, we introduce a Multi-Agent framework based on the Big Seven Personality and Socratic guidance (MAPS). This framework employs seven distinct agents that leverage feedback mechanisms and the Socratic method to guide the resolution of MSPs. To tackle the first issue, we propose a progressive four-agent solving strategy, where each agent focuses on a specific stage of the problem-solving process. For the second issue, we introduce a Critic agent, inspired by Socratic questioning, which prompts critical thinking and stimulates autonomous learning. We conduct extensive experiments on the EMMA, Olympiad, and MathVista datasets, achieving promising results that outperform the current SOTA model by 15.84% across all tasks. Meanwhile, the additional analytical experiments also verify the model's progress as well as generalization ability. |
|
2025-03-24T00:00:00 | 2503.16874 | MARS: A Multi-Agent Framework Incorporating Socratic Guidance for Automated Prompt Optimization | [
"Jian Zhang",
"Zhangqi Wang",
"Haiping Zhu",
"Jun Liu",
"Qika Lin",
"Erik Cambria"
] | The basic question-answering format of large language models involves inputting a prompt and receiving a response, and the quality of the prompt directly impacts the effectiveness of the response. Automated Prompt Optimization (APO) aims to break free from the cognitive biases of manually designed prompts and explores a broader design space for prompts. However, existing APO methods suffer from limited flexibility of fixed templates and inefficient search in prompt spaces as key issues. To this end, we propose a Multi-Agent framework Incorporating Socratic guidance (MARS), which utilizes multi-agent fusion technology for automatic planning, with gradual continuous optimization and evaluation. Specifically, MARS comprises seven agents, each with distinct functionalities, which autonomously use the Planner to devise an optimization path that ensures flexibility. Additionally, it employs a Teacher-Critic-Student Socratic dialogue pattern to iteratively optimize the prompts while conducting effective search. We conduct extensive experiments on various datasets to validate the effectiveness of our method, and perform additional analytical experiments to assess the model's advancement as well as the interpretability. |
|
2025-03-24T00:00:00 | 2503.16430 | Bridging Continuous and Discrete Tokens for Autoregressive Visual Generation | [
"Yuqing Wang",
"Zhijie Lin",
"Yao Teng",
"Yuanzhi Zhu",
"Shuhuai Ren",
"Jiashi Feng",
"Xihui Liu"
] | Autoregressive visual generation models typically rely on tokenizers to compress images into tokens that can be predicted sequentially. A fundamental dilemma exists in token representation: discrete tokens enable straightforward modeling with standard cross-entropy loss, but suffer from information loss and tokenizer training instability; continuous tokens better preserve visual details, but require complex distribution modeling, complicating the generation pipeline. In this paper, we propose TokenBridge, which bridges this gap by maintaining the strong representation capacity of continuous tokens while preserving the modeling simplicity of discrete tokens. To achieve this, we decouple discretization from the tokenizer training process through post-training quantization that directly obtains discrete tokens from continuous representations. Specifically, we introduce a dimension-wise quantization strategy that independently discretizes each feature dimension, paired with a lightweight autoregressive prediction mechanism that efficiently model the resulting large token space. Extensive experiments show that our approach achieves reconstruction and generation quality on par with continuous methods while using standard categorical prediction. This work demonstrates that bridging discrete and continuous paradigms can effectively harness the strengths of both approaches, providing a promising direction for high-quality visual generation with simple autoregressive modeling. Project page: https://yuqingwang1029.github.io/TokenBridge. |
|
2025-03-24T00:00:00 | 2503.17352 | OpenVLThinker: An Early Exploration to Complex Vision-Language Reasoning via Iterative Self-Improvement | [
"Yihe Deng",
"Hritik Bansal",
"Fan Yin",
"Nanyun Peng",
"Wei Wang",
"Kai-Wei Chang"
] | https://github.com/yihedeng9/OpenVLThinker | Recent advancements demonstrated by DeepSeek-R1 have shown that complex reasoning abilities in large language models (LLMs), including sophisticated behaviors such as self-verification and self-correction, can be achieved by RL with verifiable rewards and significantly improves model performance on challenging tasks such as AIME. Motivated by these findings, our study investigates whether similar reasoning capabilities can be successfully integrated into large vision-language models (LVLMs) and assesses their impact on challenging multimodal reasoning tasks. We consider an approach that iteratively leverages supervised fine-tuning (SFT) on lightweight training data and Reinforcement Learning (RL) to further improve model generalization. Initially, reasoning capabilities were distilled from pure-text R1 models by generating reasoning steps using high-quality captions of the images sourced from diverse visual datasets. Subsequently, iterative RL training further enhance reasoning skills, with each iteration's RL-improved model generating refined SFT datasets for the next round. This iterative process yielded OpenVLThinker, a LVLM exhibiting consistently improved reasoning performance on challenging benchmarks such as MathVista, MathVerse, and MathVision, demonstrating the potential of our strategy for robust vision-language reasoning. The code, model and data are held at https://github.com/yihedeng9/OpenVLThinker. |
2025-03-24T00:00:00 | 2503.17126 | Modifying Large Language Model Post-Training for Diverse Creative Writing | [
"John Joon Young Chung",
"Vishakh Padmakumar",
"Melissa Roemmele",
"Yuqian Sun",
"Max Kreminski"
] | As creative writing tasks do not have singular correct answers, large language models (LLMs) trained to perform these tasks should be able to generate diverse valid outputs. However, LLM post-training often focuses on improving generation quality but neglects to facilitate output diversity. Hence, in creative writing generation, we investigate post-training approaches to promote both output diversity and quality. Our core idea is to include deviation -- the degree of difference between a training sample and all other samples with the same prompt -- in the training objective to facilitate learning from rare high-quality instances. By adopting our approach to direct preference optimization (DPO) and odds ratio preference optimization (ORPO), we demonstrate that we can promote the output diversity of trained models while minimally decreasing quality. Our best model with 8B parameters could achieve on-par diversity as a human-created dataset while having output quality similar to the best instruction-tuned models we examined, GPT-4o and DeepSeek-R1. We further validate our approaches with a human evaluation, an ablation, and a comparison to an existing diversification approach, DivPO. |
|
2025-03-24T00:00:00 | 2503.17287 | FastCuRL: Curriculum Reinforcement Learning with Progressive Context Extension for Efficient Training R1-like Reasoning Models | [
"Mingyang Song",
"Mao Zheng",
"Zheng Li",
"Wenjie Yang",
"Xuan Luo",
"Yue Pan",
"Feng Zhang"
] | In this paper, we propose \textsc{FastCuRL}, a simple yet efficient Curriculum Reinforcement Learning approach with context window extending strategy to accelerate the reinforcement learning training efficiency for R1-like reasoning models while enhancing their performance in tackling complex reasoning tasks with long chain-of-thought rationales, particularly with a 1.5B parameter language model. \textsc{FastCuRL} consists of two main procedures: length-aware training data segmentation and context window extension training. Specifically, the former first splits the original training data into three different levels by the input prompt length, and then the latter leverages segmented training datasets with a progressively increasing context window length to train the reasoning model. Experimental results demonstrate that \textsc{FastCuRL}-1.5B-Preview surpasses DeepScaleR-1.5B-Preview across all five datasets (including MATH 500, AIME 2024, AMC 2023, Minerva Math, and OlympiadBench) while only utilizing 50\% of training steps. Furthermore, all training stages for FastCuRL-1.5B-Preview are completed using just a single node with 8 GPUs. |
|
2025-03-24T00:00:00 | 2503.16983 | Enabling Versatile Controls for Video Diffusion Models | [
"Xu Zhang",
"Hao Zhou",
"Haoming Qin",
"Xiaobin Lu",
"Jiaxing Yan",
"Guanzhong Wang",
"Zeyu Chen",
"Yi Liu"
] | Despite substantial progress in text-to-video generation, achieving precise and flexible control over fine-grained spatiotemporal attributes remains a significant unresolved challenge in video generation research. To address these limitations, we introduce VCtrl (also termed PP-VCtrl), a novel framework designed to enable fine-grained control over pre-trained video diffusion models in a unified manner. VCtrl integrates diverse user-specified control signals-such as Canny edges, segmentation masks, and human keypoints-into pretrained video diffusion models via a generalizable conditional module capable of uniformly encoding multiple types of auxiliary signals without modifying the underlying generator. Additionally, we design a unified control signal encoding pipeline and a sparse residual connection mechanism to efficiently incorporate control representations. Comprehensive experiments and human evaluations demonstrate that VCtrl effectively enhances controllability and generation quality. The source code and pre-trained models are publicly available and implemented using the PaddlePaddle framework at http://github.com/PaddlePaddle/PaddleMIX/tree/develop/ppdiffusers/examples/ppvctrl. |
|
2025-03-24T00:00:00 | 2503.16921 | When Preferences Diverge: Aligning Diffusion Models with Minority-Aware Adaptive DPO | [
"Lingfan Zhang",
"Chen Liu",
"Chengming Xu",
"Kai Hu",
"Donghao Luo",
"Chengjie Wang",
"Yanwei Fu",
"Yuan Yao"
] | In recent years, the field of image generation has witnessed significant advancements, particularly in fine-tuning methods that align models with universal human preferences. This paper explores the critical role of preference data in the training process of diffusion models, particularly in the context of Diffusion-DPO and its subsequent adaptations. We investigate the complexities surrounding universal human preferences in image generation, highlighting the subjective nature of these preferences and the challenges posed by minority samples in preference datasets. Through pilot experiments, we demonstrate the existence of minority samples and their detrimental effects on model performance. We propose Adaptive-DPO -- a novel approach that incorporates a minority-instance-aware metric into the DPO objective. This metric, which includes intra-annotator confidence and inter-annotator stability, distinguishes between majority and minority samples. We introduce an Adaptive-DPO loss function which improves the DPO loss in two ways: enhancing the model's learning of majority labels while mitigating the negative impact of minority samples. Our experiments demonstrate that this method effectively handles both synthetic minority data and real-world preference data, paving the way for more effective training methodologies in image generation tasks. |
|
2025-03-24T00:00:00 | 2503.16408 | RoboFactory: Exploring Embodied Agent Collaboration with Compositional Constraints | [
"Yiran Qin",
"Li Kang",
"Xiufeng Song",
"Zhenfei Yin",
"Xiaohong Liu",
"Xihui Liu",
"Ruimao Zhang",
"Lei Bai"
] | Designing effective embodied multi-agent systems is critical for solving complex real-world tasks across domains. Due to the complexity of multi-agent embodied systems, existing methods fail to automatically generate safe and efficient training data for such systems. To this end, we propose the concept of compositional constraints for embodied multi-agent systems, addressing the challenges arising from collaboration among embodied agents. We design various interfaces tailored to different types of constraints, enabling seamless interaction with the physical world. Leveraging compositional constraints and specifically designed interfaces, we develop an automated data collection framework for embodied multi-agent systems and introduce the first benchmark for embodied multi-agent manipulation, RoboFactory. Based on RoboFactory benchmark, we adapt and evaluate the method of imitation learning and analyzed its performance in different difficulty agent tasks. Furthermore, we explore the architectures and training strategies for multi-agent imitation learning, aiming to build safe and efficient embodied multi-agent systems. |
|
2025-03-24T00:00:00 | 2503.16549 | MathFlow: Enhancing the Perceptual Flow of MLLMs for Visual Mathematical Problems | [
"Felix Chen",
"Hangjie Yuan",
"Yunqiu Xu",
"Tao Feng",
"Jun Cen",
"Pengwei Liu",
"Zeying Huang",
"Yi Yang"
] | https://github.com/MathFlow-zju/MathFlow | Despite impressive performance across diverse tasks, Multimodal Large Language Models (MLLMs) have yet to fully demonstrate their potential in visual mathematical problem-solving, particularly in accurately perceiving and interpreting diagrams. Inspired by typical processes of humans, we hypothesize that the perception capabilities to extract meaningful information from diagrams is crucial, as it directly impacts subsequent inference processes. To validate this hypothesis, we developed FlowVerse, a comprehensive benchmark that categorizes all information used during problem-solving into four components, which are then combined into six problem versions for evaluation. Our preliminary results on FlowVerse reveal that existing MLLMs exhibit substantial limitations when extracting essential information and reasoned property from diagrams and performing complex reasoning based on these visual inputs. In response, we introduce MathFlow, a modular problem-solving pipeline that decouples perception and inference into distinct stages, thereby optimizing each independently. Given the perceptual limitations observed in current MLLMs, we trained MathFlow-P-7B as a dedicated perception model. Experimental results indicate that MathFlow-P-7B yields substantial performance gains when integrated with various closed-source and open-source inference models. This demonstrates the effectiveness of the MathFlow pipeline and its compatibility to diverse inference frameworks. The FlowVerse benchmark and code are available at https://github.com/MathFlow-zju/MathFlow. |
2025-03-24T00:00:00 | 2503.11572 | Implicit Bias-Like Patterns in Reasoning Models | [
"Messi H. J. Lee",
"Calvin K. Lai"
] | Implicit bias refers to automatic or spontaneous mental processes that shape perceptions, judgments, and behaviors. Previous research examining `implicit bias' in large language models (LLMs) has often approached the phenomenon differently than how it is studied in humans by focusing primarily on model outputs rather than on model processing. To examine model processing, we present a method called the Reasoning Model Implicit Association Test (RM-IAT) for studying implicit bias-like patterns in reasoning models: LLMs that employ step-by-step reasoning to solve complex tasks. Using this method, we find that reasoning models require more tokens when processing association-incompatible information compared to association-compatible information. These findings suggest AI systems harbor patterns in processing information that are analogous to human implicit bias. We consider the implications of these implicit bias-like patterns for their deployment in real-world applications. |
|
2025-03-24T00:00:00 | 2503.12821 | From Head to Tail: Towards Balanced Representation in Large Vision-Language Models through Adaptive Data Calibration | [
"Mingyang Song",
"Xiaoye Qu",
"Jiawei Zhou",
"Yu Cheng"
] | Large Vision-Language Models (LVLMs) have achieved significant progress in combining visual comprehension with language generation. Despite this success, the training data of LVLMs still suffers from Long-Tail (LT) problems, where the data distribution is highly imbalanced. Previous works have mainly focused on traditional VLM architectures, i.e., CLIP or ViT, and specific tasks such as recognition and classification. Nevertheless, the exploration of LVLM (e.g. LLaVA) and more general tasks (e.g. Visual Question Answering and Visual Reasoning) remains under-explored. In this paper, we first conduct an in-depth analysis of the LT issues in LVLMs and identify two core causes: the overrepresentation of head concepts and the underrepresentation of tail concepts. Based on the above observation, we propose an Adaptive Data Refinement Framework (ADR), which consists of two stages: Data Rebalancing (DR) and Data Synthesis (DS). In the DR stage, we adaptively rebalance the redundant data based on entity distributions, while in the DS stage, we leverage Denoising Diffusion Probabilistic Models (DDPMs) and scarce images to supplement underrepresented portions. Through comprehensive evaluations across eleven benchmarks, our proposed ADR effectively mitigates the long-tail problem in the training data, improving the average performance of LLaVA 1.5 relatively by 4.36%, without increasing the training data volume. |
|
2025-03-24T00:00:00 | 2503.16660 | When Less is Enough: Adaptive Token Reduction for Efficient Image Representation | [
"Eduard Allakhverdov",
"Elizaveta Goncharova",
"Andrey Kuznetsov"
] | Vision encoders typically generate a large number of visual tokens, providing information-rich representations but significantly increasing computational demands. This raises the question of whether all generated tokens are equally valuable or if some of them can be discarded to reduce computational costs without compromising quality. In this paper, we introduce a new method for determining feature utility based on the idea that less valuable features can be reconstructed from more valuable ones. We implement this concept by integrating an autoencoder with a Gumbel-Softmax selection mechanism, that allows identifying and retaining only the most informative visual tokens. To validate our approach, we compared the performance of the LLaVA-NeXT model, using features selected by our method with randomly selected features. We found that on OCR-based tasks, more than 50% of the visual context can be removed with minimal performance loss, whereas randomly discarding the same proportion of features significantly affects the model capabilities. Furthermore, in general-domain tasks, even randomly retaining only 30% of tokens achieves performance comparable to using the full set of visual tokens. Our results highlight a promising direction towards adaptive and efficient multimodal pruning that facilitates scalable and low-overhead inference without compromising performance. |
|
2025-03-24T00:00:00 | 2503.16867 | ETVA: Evaluation of Text-to-Video Alignment via Fine-grained Question Generation and Answering | [
"Kaisi Guan",
"Zhengfeng Lai",
"Yuchong Sun",
"Peng Zhang",
"Wei Liu",
"Kieran Liu",
"Meng Cao",
"Ruihua Song"
] | Precisely evaluating semantic alignment between text prompts and generated videos remains a challenge in Text-to-Video (T2V) Generation. Existing text-to-video alignment metrics like CLIPScore only generate coarse-grained scores without fine-grained alignment details, failing to align with human preference. To address this limitation, we propose ETVA, a novel Evaluation method of Text-to-Video Alignment via fine-grained question generation and answering. First, a multi-agent system parses prompts into semantic scene graphs to generate atomic questions. Then we design a knowledge-augmented multi-stage reasoning framework for question answering, where an auxiliary LLM first retrieves relevant common-sense knowledge (e.g., physical laws), and then video LLM answers the generated questions through a multi-stage reasoning mechanism. Extensive experiments demonstrate that ETVA achieves a Spearman's correlation coefficient of 58.47, showing a much higher correlation with human judgment than existing metrics which attain only 31.0. We also construct a comprehensive benchmark specifically designed for text-to-video alignment evaluation, featuring 2k diverse prompts and 12k atomic questions spanning 10 categories. Through a systematic evaluation of 15 existing text-to-video models, we identify their key capabilities and limitations, paving the way for next-generation T2V generation. |
|
2025-03-24T00:00:00 | 2503.17069 | PVChat: Personalized Video Chat with One-Shot Learning | [
"Yufei Shi",
"Weilong Yan",
"Gang Xu",
"Yumeng Li",
"Yuchen Li",
"Zhenxi Li",
"Fei Richard Yu",
"Ming Li",
"Si Yong Yeo"
] | Video large language models (ViLLMs) excel in general video understanding, e.g., recognizing activities like talking and eating, but struggle with identity-aware comprehension, such as "Wilson is receiving chemotherapy" or "Tom is discussing with Sarah", limiting their applicability in smart healthcare and smart home environments. To address this limitation, we propose a one-shot learning framework PVChat, the first personalized ViLLM that enables subject-aware question answering (QA) from a single video for each subject. Our approach optimizes a Mixture-of-Heads (MoH) enhanced ViLLM on a synthetically augmented video-QA dataset, leveraging a progressive image-to-video learning strategy. Specifically, we introduce an automated augmentation pipeline that synthesizes identity-preserving positive samples and retrieves hard negatives from existing video corpora, generating a diverse training dataset with four QA types: existence, appearance, action, and location inquiries. To enhance subject-specific learning, we propose a ReLU Routing MoH attention mechanism, alongside two novel objectives: (1) Smooth Proximity Regularization for progressive learning through exponential distance scaling and (2) Head Activation Enhancement for balanced attention routing. Finally, we adopt a two-stage training strategy, transitioning from image pre-training to video fine-tuning, enabling a gradual learning process from static attributes to dynamic representations. We evaluate PVChat on diverse datasets covering medical scenarios, TV series, anime, and real-world footage, demonstrating its superiority in personalized feature understanding after learning from a single video, compared to state-of-the-art ViLLMs. |
|
2025-03-24T00:00:00 | 2503.16025 | Single Image Iterative Subject-driven Generation and Editing | [
"Yair Shpitzer",
"Gal Chechik",
"Idan Schwartz"
] | Personalizing image generation and editing is particularly challenging when we only have a few images of the subject, or even a single image. A common approach to personalization is concept learning, which can integrate the subject into existing models relatively quickly, but produces images whose quality tends to deteriorate quickly when the number of subject images is small. Quality can be improved by pre-training an encoder, but training restricts generation to the training distribution, and is time consuming. It is still an open hard challenge to personalize image generation and editing from a single image without training. Here, we present SISO, a novel, training-free approach based on optimizing a similarity score with an input subject image. More specifically, SISO iteratively generates images and optimizes the model based on loss of similarity with the given subject image until a satisfactory level of similarity is achieved, allowing plug-and-play optimization to any image generator. We evaluated SISO in two tasks, image editing and image generation, using a diverse data set of personal subjects, and demonstrate significant improvements over existing methods in image quality, subject fidelity, and background preservation. |
|
2025-03-24T00:00:00 | 2503.17032 | TaoAvatar: Real-Time Lifelike Full-Body Talking Avatars for Augmented Reality via 3D Gaussian Splatting | [
"Jianchuan Chen",
"Jingchuan Hu",
"Gaige Wang",
"Zhonghua Jiang",
"Tiansong Zhou",
"Zhiwen Chen",
"Chengfei Lv"
] | Realistic 3D full-body talking avatars hold great potential in AR, with applications ranging from e-commerce live streaming to holographic communication. Despite advances in 3D Gaussian Splatting (3DGS) for lifelike avatar creation, existing methods struggle with fine-grained control of facial expressions and body movements in full-body talking tasks. Additionally, they often lack sufficient details and cannot run in real-time on mobile devices. We present TaoAvatar, a high-fidelity, lightweight, 3DGS-based full-body talking avatar driven by various signals. Our approach starts by creating a personalized clothed human parametric template that binds Gaussians to represent appearances. We then pre-train a StyleUnet-based network to handle complex pose-dependent non-rigid deformation, which can capture high-frequency appearance details but is too resource-intensive for mobile devices. To overcome this, we "bake" the non-rigid deformations into a lightweight MLP-based network using a distillation technique and develop blend shapes to compensate for details. Extensive experiments show that TaoAvatar achieves state-of-the-art rendering quality while running in real-time across various devices, maintaining 90 FPS on high-definition stereo devices such as the Apple Vision Pro. |
|
2025-03-24T00:00:00 | 2503.16282 | Generalized Few-shot 3D Point Cloud Segmentation with Vision-Language Model | [
"Zhaochong An",
"Guolei Sun",
"Yun Liu",
"Runjia Li",
"Junlin Han",
"Ender Konukoglu",
"Serge Belongie"
] | https://github.com/ZhaochongAn/GFS-VL | Generalized few-shot 3D point cloud segmentation (GFS-PCS) adapts models to new classes with few support samples while retaining base class segmentation. Existing GFS-PCS methods enhance prototypes via interacting with support or query features but remain limited by sparse knowledge from few-shot samples. Meanwhile, 3D vision-language models (3D VLMs), generalizing across open-world novel classes, contain rich but noisy novel class knowledge. In this work, we introduce a GFS-PCS framework that synergizes dense but noisy pseudo-labels from 3D VLMs with precise yet sparse few-shot samples to maximize the strengths of both, named GFS-VL. Specifically, we present a prototype-guided pseudo-label selection to filter low-quality regions, followed by an adaptive infilling strategy that combines knowledge from pseudo-label contexts and few-shot samples to adaptively label the filtered, unlabeled areas. Additionally, we design a novel-base mix strategy to embed few-shot samples into training scenes, preserving essential context for improved novel class learning. Moreover, recognizing the limited diversity in current GFS-PCS benchmarks, we introduce two challenging benchmarks with diverse novel classes for comprehensive generalization evaluation. Experiments validate the effectiveness of our framework across models and datasets. Our approach and benchmarks provide a solid foundation for advancing GFS-PCS in the real world. The code is at https://github.com/ZhaochongAn/GFS-VL |
2025-03-24T00:00:00 | 2503.17095 | FFaceNeRF: Few-shot Face Editing in Neural Radiance Fields | [
"Kwan Yun",
"Chaelin Kim",
"Hangyeul Shin",
"Junyong Noh"
] | Recent 3D face editing methods using masks have produced high-quality edited images by leveraging Neural Radiance Fields (NeRF). Despite their impressive performance, existing methods often provide limited user control due to the use of pre-trained segmentation masks. To utilize masks with a desired layout, an extensive training dataset is required, which is challenging to gather. We present FFaceNeRF, a NeRF-based face editing technique that can overcome the challenge of limited user control due to the use of fixed mask layouts. Our method employs a geometry adapter with feature injection, allowing for effective manipulation of geometry attributes. Additionally, we adopt latent mixing for tri-plane augmentation, which enables training with a few samples. This facilitates rapid model adaptation to desired mask layouts, crucial for applications in fields like personalized medical imaging or creative face editing. Our comparative evaluations demonstrate that FFaceNeRF surpasses existing mask based face editing methods in terms of flexibility, control, and generated image quality, paving the way for future advancements in customized and high-fidelity 3D face editing. The code is available on the {https://kwanyun.github.io/FFaceNeRF_page/{project-page}}. |
|
2025-03-24T00:00:00 | 2503.16423 | GAEA: A Geolocation Aware Conversational Model | [
"Ron Campos",
"Ashmal Vayani",
"Parth Parag Kulkarni",
"Rohit Gupta",
"Aritra Dutta",
"Mubarak Shah"
] | Image geolocalization, in which, traditionally, an AI model predicts the precise GPS coordinates of an image is a challenging task with many downstream applications. However, the user cannot utilize the model to further their knowledge other than the GPS coordinate; the model lacks an understanding of the location and the conversational ability to communicate with the user. In recent days, with tremendous progress of large multimodal models (LMMs) proprietary and open-source researchers have attempted to geolocalize images via LMMs. However, the issues remain unaddressed; beyond general tasks, for more specialized downstream tasks, one of which is geolocalization, LMMs struggle. In this work, we propose to solve this problem by introducing a conversational model GAEA that can provide information regarding the location of an image, as required by a user. No large-scale dataset enabling the training of such a model exists. Thus we propose a comprehensive dataset GAEA with 800K images and around 1.6M question answer pairs constructed by leveraging OpenStreetMap (OSM) attributes and geographical context clues. For quantitative evaluation, we propose a diverse benchmark comprising 4K image-text pairs to evaluate conversational capabilities equipped with diverse question types. We consider 11 state-of-the-art open-source and proprietary LMMs and demonstrate that GAEA significantly outperforms the best open-source model, LLaVA-OneVision by 25.69% and the best proprietary model, GPT-4o by 8.28%. Our dataset, model and codes are available |
|
2025-03-24T00:00:00 | 2503.14607 | Can Large Vision Language Models Read Maps Like a Human? | [
"Shuo Xing",
"Zezhou Sun",
"Shuangyu Xie",
"Kaiyuan Chen",
"Yanjia Huang",
"Yuping Wang",
"Jiachen Li",
"Dezhen Song",
"Zhengzhong Tu"
] | https://github.com/taco-group/MapBench | In this paper, we introduce MapBench-the first dataset specifically designed for human-readable, pixel-based map-based outdoor navigation, curated from complex path finding scenarios. MapBench comprises over 1600 pixel space map path finding problems from 100 diverse maps. In MapBench, LVLMs generate language-based navigation instructions given a map image and a query with beginning and end landmarks. For each map, MapBench provides Map Space Scene Graph (MSSG) as an indexing data structure to convert between natural language and evaluate LVLM-generated results. We demonstrate that MapBench significantly challenges state-of-the-art LVLMs both zero-shot prompting and a Chain-of-Thought (CoT) augmented reasoning framework that decomposes map navigation into sequential cognitive processes. Our evaluation of both open-source and closed-source LVLMs underscores the substantial difficulty posed by MapBench, revealing critical limitations in their spatial reasoning and structured decision-making capabilities. We release all the code and dataset in https://github.com/taco-group/MapBench. |
2025-03-24T00:00:00 | 2503.17407 | A Comprehensive Survey on Long Context Language Modeling | [
"Jiaheng Liu",
"Dawei Zhu",
"Zhiqi Bai",
"Yancheng He",
"Huanxuan Liao",
"Haoran Que",
"Zekun Wang",
"Chenchen Zhang",
"Ge Zhang",
"Jiebin Zhang",
"Yuanxing Zhang",
"Zhuo Chen",
"Hangyu Guo",
"Shilong Li",
"Ziqiang Liu",
"Yong Shan",
"Yifan Song",
"Jiayi Tian",
"Wenhao Wu",
"Zhejian Zhou",
"Ruijie Zhu",
"Junlan Feng",
"Yang Gao",
"Shizhu He",
"Zhoujun Li",
"Tianyu Liu",
"Fanyu Meng",
"Wenbo Su",
"Yingshui Tan",
"Zili Wang",
"Jian Yang",
"Wei Ye",
"Bo Zheng",
"Wangchunshu Zhou",
"Wenhao Huang",
"Sujian Li",
"Zhaoxiang Zhang"
] | https://github.com/LCLM-Horizon/A-Comprehensive-Survey-For-Long-Context-Language-Modeling{\color[RGB]{175 | Efficient processing of long contexts has been a persistent pursuit in Natural Language Processing. With the growing number of long documents, dialogues, and other textual data, it is important to develop Long Context Language Models (LCLMs) that can process and analyze extensive inputs in an effective and efficient way. In this paper, we present a comprehensive survey on recent advances in long-context modeling for large language models. Our survey is structured around three key aspects: how to obtain effective and efficient LCLMs, how to train and deploy LCLMs efficiently, and how to evaluate and analyze LCLMs comprehensively. For the first aspect, we discuss data strategies, architectural designs, and workflow approaches oriented with long context processing. For the second aspect, we provide a detailed examination of the infrastructure required for LCLM training and inference. For the third aspect, we present evaluation paradigms for long-context comprehension and long-form generation, as well as behavioral analysis and mechanism interpretability of LCLMs. Beyond these three key aspects, we thoroughly explore the diverse application scenarios where existing LCLMs have been deployed and outline promising future development directions. This survey provides an up-to-date review of the literature on long-context LLMs, which we wish to serve as a valuable resource for both researchers and engineers. An associated GitHub repository collecting the latest papers and repos is available at: https://github.com/LCLM-Horizon/A-Comprehensive-Survey-For-Long-Context-Language-Modeling{\color[RGB]{175,36,67}{LCLM-Horizon}}. |
2025-03-25T00:00:00 | 2503.18769 | AlphaSpace: Enabling Robotic Actions through Semantic Tokenization and Symbolic Reasoning | [
"Alan Dao",
"Dinh Bach Vu",
"Bui Quang Huy"
] | This paper presents AlphaSpace, a novel methodology designed to enhance the spatial reasoning capabilities of large language models (LLMs) for 3D Cartesian space navigation. AlphaSpace employs a semantics-based tokenization strategy, encoding height information through specialized semantic tokens, and integrates primarily symbolic synthetic reasoning data. This approach enables LLMs to accurately manipulate objects by positioning them at specific [x, y, z] coordinates. Experimental results demonstrate that AlphaSpace significantly outperforms existing models on manipulation subtasks, achieving a total accuracy of 66.67%, compared to 37.5% for GPT-4o and 29.17% for Claude 3.5 Sonnet. |
|
2025-03-25T00:00:00 | 2503.14428 | MagicComp: Training-free Dual-Phase Refinement for Compositional Video Generation | [
"Hongyu Zhang",
"Yufan Deng",
"Shenghai Yuan",
"Peng Jin",
"Zesen Cheng",
"Yian Zhao",
"Chang Liu",
"Jie Chen"
] | Text-to-video (T2V) generation has made significant strides with diffusion models. However, existing methods still struggle with accurately binding attributes, determining spatial relationships, and capturing complex action interactions between multiple subjects. To address these limitations, we propose MagicComp, a training-free method that enhances compositional T2V generation through dual-phase refinement. Specifically, (1) During the Conditioning Stage: We introduce the Semantic Anchor Disambiguation to reinforces subject-specific semantics and resolve inter-subject ambiguity by progressively injecting the directional vectors of semantic anchors into original text embedding; (2) During the Denoising Stage: We propose Dynamic Layout Fusion Attention, which integrates grounding priors and model-adaptive spatial perception to flexibly bind subjects to their spatiotemporal regions through masked attention modulation. Furthermore, MagicComp is a model-agnostic and versatile approach, which can be seamlessly integrated into existing T2V architectures. Extensive experiments on T2V-CompBench and VBench demonstrate that MagicComp outperforms state-of-the-art methods, highlighting its potential for applications such as complex prompt-based and trajectory-controllable video generation. Project page: https://hong-yu-zhang.github.io/MagicComp-Page/. |
|
2025-03-25T00:00:00 | 2503.17359 | Position: Interactive Generative Video as Next-Generation Game Engine | [
"Jiwen Yu",
"Yiran Qin",
"Haoxuan Che",
"Quande Liu",
"Xintao Wang",
"Pengfei Wan",
"Di Zhang",
"Xihui Liu"
] | Modern game development faces significant challenges in creativity and cost due to predetermined content in traditional game engines. Recent breakthroughs in video generation models, capable of synthesizing realistic and interactive virtual environments, present an opportunity to revolutionize game creation. In this position paper, we propose Interactive Generative Video (IGV) as the foundation for Generative Game Engines (GGE), enabling unlimited novel content generation in next-generation gaming. GGE leverages IGV's unique strengths in unlimited high-quality content synthesis, physics-aware world modeling, user-controlled interactivity, long-term memory capabilities, and causal reasoning. We present a comprehensive framework detailing GGE's core modules and a hierarchical maturity roadmap (L0-L4) to guide its evolution. Our work charts a new course for game development in the AI era, envisioning a future where AI-powered generative systems fundamentally reshape how games are created and experienced. |
|
2025-03-25T00:00:00 | 2503.18942 | Video-T1: Test-Time Scaling for Video Generation | [
"Fangfu Liu",
"Hanyang Wang",
"Yimo Cai",
"Kaiyan Zhang",
"Xiaohang Zhan",
"Yueqi Duan"
] | With the scale capability of increasing training data, model size, and computational cost, video generation has achieved impressive results in digital creation, enabling users to express creativity across various domains. Recently, researchers in Large Language Models (LLMs) have expanded the scaling to test-time, which can significantly improve LLM performance by using more inference-time computation. Instead of scaling up video foundation models through expensive training costs, we explore the power of Test-Time Scaling (TTS) in video generation, aiming to answer the question: if a video generation model is allowed to use non-trivial amount of inference-time compute, how much can it improve generation quality given a challenging text prompt. In this work, we reinterpret the test-time scaling of video generation as a searching problem to sample better trajectories from Gaussian noise space to the target video distribution. Specifically, we build the search space with test-time verifiers to provide feedback and heuristic algorithms to guide searching process. Given a text prompt, we first explore an intuitive linear search strategy by increasing noise candidates at inference time. As full-step denoising all frames simultaneously requires heavy test-time computation costs, we further design a more efficient TTS method for video generation called Tree-of-Frames (ToF) that adaptively expands and prunes video branches in an autoregressive manner. Extensive experiments on text-conditioned video generation benchmarks demonstrate that increasing test-time compute consistently leads to significant improvements in the quality of videos. Project page: https://liuff19.github.io/Video-T1 |
|
2025-03-25T00:00:00 | 2503.18940 | Training-free Diffusion Acceleration with Bottleneck Sampling | [
"Ye Tian",
"Xin Xia",
"Yuxi Ren",
"Shanchuan Lin",
"Xing Wang",
"Xuefeng Xiao",
"Yunhai Tong",
"Ling Yang",
"Bin Cui"
] | https://github.com/tyfeld/Bottleneck-Sampling | Diffusion models have demonstrated remarkable capabilities in visual content generation but remain challenging to deploy due to their high computational cost during inference. This computational burden primarily arises from the quadratic complexity of self-attention with respect to image or video resolution. While existing acceleration methods often compromise output quality or necessitate costly retraining, we observe that most diffusion models are pre-trained at lower resolutions, presenting an opportunity to exploit these low-resolution priors for more efficient inference without degrading performance. In this work, we introduce Bottleneck Sampling, a training-free framework that leverages low-resolution priors to reduce computational overhead while preserving output fidelity. Bottleneck Sampling follows a high-low-high denoising workflow: it performs high-resolution denoising in the initial and final stages while operating at lower resolutions in intermediate steps. To mitigate aliasing and blurring artifacts, we further refine the resolution transition points and adaptively shift the denoising timesteps at each stage. We evaluate Bottleneck Sampling on both image and video generation tasks, where extensive experiments demonstrate that it accelerates inference by up to 3times for image generation and 2.5times for video generation, all while maintaining output quality comparable to the standard full-resolution sampling process across multiple evaluation metrics. Code is available at: https://github.com/tyfeld/Bottleneck-Sampling |
2025-03-25T00:00:00 | 2503.17489 | Judge Anything: MLLM as a Judge Across Any Modality | [
"Shu Pu",
"Yaochen Wang",
"Dongping Chen",
"Yuhang Chen",
"Guohao Wang",
"Qi Qin",
"Zhongyi Zhang",
"Zhiyuan Zhang",
"Zetong Zhou",
"Shuang Gong",
"Yi Gui",
"Yao Wan",
"Philip S. Yu"
] | Evaluating generative foundation models on open-ended multimodal understanding (MMU) and generation (MMG) tasks across diverse modalities (e.g., images, audio, video) poses significant challenges due to the complexity of cross-modal interactions. To this end, the idea of utilizing Multimodal LLMs (MLLMs) as automated judges has emerged, with encouraging results in assessing vision-language understanding tasks. Moving further, this paper extends MLLM-as-a-Judge across modalities to a unified manner by introducing two benchmarks, TaskAnything and JudgeAnything, to respectively evaluate the overall performance and judging capabilities of MLLMs across any-to-any modality tasks. Specifically, TaskAnything evaluates the MMU and MMG capabilities across 15 any-to-any modality categories, employing 1,500 queries curated from well-established benchmarks. Furthermore, JudgeAnything evaluates the judging capabilities of 5 advanced (e.g., GPT-4o and Gemini-2.0-Flash) from the perspectives of Pair Comparison and Score Evaluation, providing a standardized testbed that incorporates human judgments and detailed rubrics. Our extensive experiments reveal that while these MLLMs show promise in assessing MMU (i.e., achieving an average of 66.55% in Pair Comparison setting and 42.79% in Score Evaluation setting), they encounter significant challenges with MMG tasks (i.e., averaging only 53.37% in Pair Comparison setting and 30.05% in Score Evaluation setting), exposing cross-modality biases and hallucination issues. To address this, we present OmniArena, an automated platform for evaluating omni-models and multimodal reward models. Our work highlights the need for fairer evaluation protocols and stronger alignment with human preferences. The source code and dataset are publicly available at: https://urrealhero.github.io/judgeanythingweb/. |
|
2025-03-25T00:00:00 | 2503.17439 | LEMMA: Learning from Errors for MatheMatical Advancement in LLMs | [
"Zhuoshi Pan",
"Yu Li",
"Honglin Lin",
"Qizhi Pei",
"Zinan Tang",
"Wei Wu",
"Chenlin Ming",
"H. Vicky Zhao",
"Conghui He",
"Lijun Wu"
] | Large language models (LLMs) have demonstrated remarkable reasoning capability in solving mathematical problems. However, existing approaches primarily focus on improving the quality of correct training data, e.g., distilling high-quality correct solutions from advanced models, neglecting the value contained in error data, potentially hindering the model's reflective ability. Though some studies attempt to leverage error data, they often involve complex mechanisms, such as Monte Carlo Tree Search (MCTS) to explore error nodes. In this work, we propose to enhance LLMs' reasoning ability by Learning from Errors for Mathematical Advancement (LEMMA). LEMMA constructs data consisting of an incorrect solution with an erroneous step and a reflection connection to a correct solution for fine-tuning. Specifically, we systematically analyze the model-generated error types and introduce an error-type grounded mistake augmentation method to collect diverse and representative errors. Correct solutions are either from fixing the errors or generating a fresh start. Through a model-aware smooth reflection connection, the erroneous solution is transferred to the correct one. By fine-tuning on the constructed dataset, the model is able to self-correct errors autonomously within the generation process without relying on external critique models. Experimental results demonstrate that LEMMA achieves significant performance improvements over other strong baselines. |
|
2025-03-25T00:00:00 | 2503.18923 | Video SimpleQA: Towards Factuality Evaluation in Large Video Language Models | [
"Meng Cao",
"Pengfei Hu",
"Yingyao Wang",
"Jihao Gu",
"Haoran Tang",
"Haoze Zhao",
"Jiahua Dong",
"Wangbo Yu",
"Ge Zhang",
"Ian Reid",
"Xiaodan Liang"
] | Recent advancements in Large Video Language Models (LVLMs) have highlighted their potential for multi-modal understanding, yet evaluating their factual grounding in video contexts remains a critical unsolved challenge. To address this gap, we introduce Video SimpleQA, the first comprehensive benchmark tailored for factuality evaluation of LVLMs. Our work distinguishes from existing video benchmarks through the following key features: 1) Knowledge required: demanding integration of external knowledge beyond the explicit narrative; 2) Fact-seeking question: targeting objective, undisputed events or relationships, avoiding subjective interpretation; 3) Definitive & short-form answer: Answers are crafted as unambiguous and definitively correct in a short format, enabling automated evaluation through LLM-as-a-judge frameworks with minimal scoring variance; 4) External-source verified: All annotations undergo rigorous validation against authoritative external references to ensure the reliability; 5) Temporal reasoning required: The annotated question types encompass both static single-frame understanding and dynamic temporal reasoning, explicitly evaluating LVLMs factuality under the long-context dependencies. We extensively evaluate 41 state-of-the-art LVLMs and summarize key findings as follows: 1) Current LVLMs exhibit notable deficiencies in factual adherence, particularly for open-source models. The best-performing model Gemini-1.5-Pro achieves merely an F-score of 54.4%; 2) Test-time compute paradigms show insignificant performance gains, revealing fundamental constraints for enhancing factuality through post-hoc computation; 3) Retrieval-Augmented Generation demonstrates consistent improvements at the cost of additional inference time overhead, presenting a critical efficiency-performance trade-off. |
|
2025-03-25T00:00:00 | 2503.17422 | V-Seek: Accelerating LLM Reasoning on Open-hardware Server-class RISC-V Platforms | [
"Javier J. Poveda Rodrigo",
"Mohamed Amine Ahmdi",
"Alessio Burrello",
"Daniele Jahier Pagliari",
"Luca Benini"
] | The recent exponential growth of Large Language Models (LLMs) has relied on GPU-based systems. However, CPUs are emerging as a flexible and lower-cost alternative, especially when targeting inference and reasoning workloads. RISC-V is rapidly gaining traction in this area, given its open and vendor-neutral ISA. However, the RISC-V hardware for LLM workloads and the corresponding software ecosystem are not fully mature and streamlined, given the requirement of domain-specific tuning. This paper aims at filling this gap, focusing on optimizing LLM inference on the Sophon SG2042, the first commercially available many-core RISC-V CPU with vector processing capabilities. On two recent state-of-the-art LLMs optimized for reasoning, DeepSeek R1 Distill Llama 8B and DeepSeek R1 Distill QWEN 14B, we achieve 4.32/2.29 token/s for token generation and 6.54/3.68 token/s for prompt processing, with a speed up of up 2.9x/3.0x compared to our baseline. |
|
2025-03-25T00:00:00 | 2503.18892 | SimpleRL-Zoo: Investigating and Taming Zero Reinforcement Learning for Open Base Models in the Wild | [
"Weihao Zeng",
"Yuzhen Huang",
"Qian Liu",
"Wei Liu",
"Keqing He",
"Zejun Ma",
"Junxian He"
] | DeepSeek-R1 has shown that long chain-of-thought (CoT) reasoning can naturally emerge through a simple reinforcement learning (RL) framework with rule-based rewards, where the training may directly start from the base models-a paradigm referred to as zero RL training. Most recent efforts to reproduce zero RL training have primarily focused on the Qwen2.5 model series, which may not be representative as we find the base models already exhibit strong instruction-following and self-reflection abilities. In this work, we investigate zero RL training across 10 diverse base models, spanning different families and sizes including LLama3-8B, Mistral-7B/24B, DeepSeek-Math-7B, Qwen2.5-math-7B, and all Qwen2.5 models from 0.5B to 32B. Leveraging several key design strategies-such as adjusting format reward and controlling query difficulty-we achieve substantial improvements in both reasoning accuracy and response length across most settings. However, by carefully monitoring the training dynamics, we observe that different base models exhibit distinct patterns during training. For instance, the increased response length does not always correlate with the emergence of certain cognitive behaviors such as verification (i.e., the "aha moment"). Notably, we observe the "aha moment" for the first time in small models not from the Qwen family. We share the key designs that enable successful zero RL training, along with our findings and practices. To facilitate further research, we open-source the code, models, and analysis tools. |
|
2025-03-25T00:00:00 | 2503.18013 | Vision-R1: Evolving Human-Free Alignment in Large Vision-Language Models via Vision-Guided Reinforcement Learning | [
"Yufei Zhan",
"Yousong Zhu",
"Shurong Zheng",
"Hongyin Zhao",
"Fan Yang",
"Ming Tang",
"Jinqiao Wang"
] | Large Vision-Language Models (LVLMs) typically follow a two-stage training paradigm-pretraining and supervised fine-tuning. Recently, preference optimization, derived from the language domain, has emerged as an effective post-training reinforcement strategy to enhance capabilities of LVLMs. However, constructing high-quality human-annotated preference data and developing robust reward models to mimic these preferences are both costly and challenging. Motivated by this observation, we propose Vision-R1, a novel vision-guided R1-like reinforcement learning algorithm for LVLMs that rewards models with definitive vision feedback. It only leverages curated instruction data, eliminating the need for specialized reward models and handcrafted preference datasets. We incorporate a criterion-driven reward function that further integrates multi-dimensional feedback to evaluate model completions comprehensively based on the vision task logic. Furthermore, we introduce a progressive rule refinement strategy that dynamically adjusts the reward criteria during training, enabling continuous model improvement and mitigating reward hacking. Extensive experiments on both in-distribution and out-of-distribution benchmarks demonstrate that fine-tuning the 7B LVLMs with Vision-R1 achieves consistent performance gains, with even up to 50% improvement and surpassing the state-of-the-art 10x size model. |
|
2025-03-25T00:00:00 | 2503.17735 | RDTF: Resource-efficient Dual-mask Training Framework for Multi-frame Animated Sticker Generation | [
"Zhiqiang Yuan",
"Ting Zhang",
"Ying Deng",
"Jiapei Zhang",
"Yeshuang Zhu",
"Zexi Jia",
"Jie Zhou",
"Jinchao Zhang"
] | Recently, great progress has been made in video generation technology, attracting the widespread attention of scholars. To apply this technology to downstream applications under resource-constrained conditions, researchers usually fine-tune the pre-trained models based on parameter-efficient tuning methods such as Adapter or Lora. Although these methods can transfer the knowledge from the source domain to the target domain, fewer training parameters lead to poor fitting ability, and the knowledge from the source domain may lead to the inference process deviating from the target domain. In this paper, we argue that under constrained resources, training a smaller video generation model from scratch using only million-level samples can outperform parameter-efficient tuning on larger models in downstream applications: the core lies in the effective utilization of data and curriculum strategy. Take animated sticker generation (ASG) as a case study, we first construct a discrete frame generation network for stickers with low frame rates, ensuring that its parameters meet the requirements of model training under constrained resources. In order to provide data support for models trained from scratch, we come up with a dual-mask based data utilization strategy, which manages to improve the availability and expand the diversity of limited data. To facilitate convergence under dual-mask situation, we propose a difficulty-adaptive curriculum learning method, which decomposes the sample entropy into static and adaptive components so as to obtain samples from easy to difficult. The experiment demonstrates that our resource-efficient dual-mask training framework is quantitatively and qualitatively superior to efficient-parameter tuning methods such as I2V-Adapter and SimDA, verifying the feasibility of our method on downstream tasks under constrained resources. Code will be available. |
|
2025-03-25T00:00:00 | 2503.18945 | Aether: Geometric-Aware Unified World Modeling | [
"Aether Team",
"Haoyi Zhu",
"Yifan Wang",
"Jianjun Zhou",
"Wenzheng Chang",
"Yang Zhou",
"Zizun Li",
"Junyi Chen",
"Chunhua Shen",
"Jiangmiao Pang",
"Tong He"
] | The integration of geometric reconstruction and generative modeling remains a critical challenge in developing AI systems capable of human-like spatial reasoning. This paper proposes Aether, a unified framework that enables geometry-aware reasoning in world models by jointly optimizing three core capabilities: (1) 4D dynamic reconstruction, (2) action-conditioned video prediction, and (3) goal-conditioned visual planning. Through task-interleaved feature learning, Aether achieves synergistic knowledge sharing across reconstruction, prediction, and planning objectives. Building upon video generation models, our framework demonstrates unprecedented synthetic-to-real generalization despite never observing real-world data during training. Furthermore, our approach achieves zero-shot generalization in both action following and reconstruction tasks, thanks to its intrinsic geometric modeling. Remarkably, even without real-world data, its reconstruction performance far exceeds that of domain-specific models. Additionally, Aether leverages a geometry-informed action space to seamlessly translate predictions into actions, enabling effective autonomous trajectory planning. We hope our work inspires the community to explore new frontiers in physically-reasonable world modeling and its applications. |
|
2025-03-25T00:00:00 | 2503.18908 | FFN Fusion: Rethinking Sequential Computation in Large Language Models | [
"Akhiad Bercovich",
"Mohammad Dabbah",
"Omri Puny",
"Ido Galil",
"Amnon Geifman",
"Yonatan Geifman",
"Izhak Golan",
"Ehud Karpas",
"Itay Levy",
"Zach Moshe",
"Najeeb Nabwani",
"Tomer Ronen",
"Itamar Schen",
"Elad Segal",
"Ido Shahaf",
"Oren Tropp",
"Ran Zilberstein",
"Ran El-Yaniv"
] | We introduce FFN Fusion, an architectural optimization technique that reduces sequential computation in large language models by identifying and exploiting natural opportunities for parallelization. Our key insight is that sequences of Feed-Forward Network (FFN) layers, particularly those remaining after the removal of specific attention layers, can often be parallelized with minimal accuracy impact. We develop a principled methodology for identifying and fusing such sequences, transforming them into parallel operations that significantly reduce inference latency while preserving model behavior. Applying these techniques to Llama-3.1-405B-Instruct, we create Llama-Nemotron-Ultra-253B-Base (Ultra-253B-Base), an efficient and soon-to-be publicly available model that achieves a 1.71X speedup in inference latency and 35X lower per-token cost while maintaining strong performance across benchmarks. Through extensive experiments on models from 49B to 253B parameters, we demonstrate that FFN Fusion becomes increasingly effective at larger scales and can complement existing optimization techniques like quantization and pruning. Most intriguingly, we find that even full transformer blocks containing both attention and FFN layers can sometimes be parallelized, suggesting new directions for neural architecture design. |
|
2025-03-25T00:00:00 | 2503.18866 | Reasoning to Learn from Latent Thoughts | [
"Yangjun Ruan",
"Neil Band",
"Chris J. Maddison",
"Tatsunori Hashimoto"
] | Compute scaling for language model (LM) pretraining has outpaced the growth of human-written texts, leading to concerns that data will become the bottleneck to LM scaling. To continue scaling pretraining in this data-constrained regime, we propose that explicitly modeling and inferring the latent thoughts that underlie the text generation process can significantly improve pretraining data efficiency. Intuitively, our approach views web text as the compressed final outcome of a verbose human thought process and that the latent thoughts contain important contextual knowledge and reasoning steps that are critical to data-efficient learning. We empirically demonstrate the effectiveness of our approach through data-constrained continued pretraining for math. We first show that synthetic data approaches to inferring latent thoughts significantly improve data efficiency, outperforming training on the same amount of raw data (5.7\% rightarrow 25.4\% on MATH). Furthermore, we demonstrate latent thought inference without a strong teacher, where an LM bootstraps its own performance by using an EM algorithm to iteratively improve the capability of the trained LM and the quality of thought-augmented pretraining data. We show that a 1B LM can bootstrap its performance across at least three iterations and significantly outperform baselines trained on raw data, with increasing gains from additional inference compute when performing the E-step. The gains from inference scaling and EM iterations suggest new opportunities for scaling data-constrained pretraining. |
|
2025-03-25T00:00:00 | 2503.18559 | AMD-Hummingbird: Towards an Efficient Text-to-Video Model | [
"Takashi Isobe",
"He Cui",
"Dong Zhou",
"Mengmeng Ge",
"Dong Li",
"Emad Barsoum"
] | Text-to-Video (T2V) generation has attracted significant attention for its ability to synthesize realistic videos from textual descriptions. However, existing models struggle to balance computational efficiency and high visual quality, particularly on resource-limited devices, e.g.,iGPUs and mobile phones. Most prior work prioritizes visual fidelity while overlooking the need for smaller, more efficient models suitable for real-world deployment. To address this challenge, we propose a lightweight T2V framework, termed Hummingbird, which prunes existing models and enhances visual quality through visual feedback learning. Our approach reduces the size of the U-Net from 1.4 billion to 0.7 billion parameters, significantly improving efficiency while preserving high-quality video generation. Additionally, we introduce a novel data processing pipeline that leverages Large Language Models (LLMs) and Video Quality Assessment (VQA) models to enhance the quality of both text prompts and video data. To support user-driven training and style customization, we publicly release the full training code, including data processing and model training. Extensive experiments show that our method achieves a 31X speedup compared to state-of-the-art models such as VideoCrafter2, while also attaining the highest overall score on VBench. Moreover, our method supports the generation of videos with up to 26 frames, addressing the limitations of existing U-Net-based methods in long video generation. Notably, the entire training process requires only four GPUs, yet delivers performance competitive with existing leading methods. Hummingbird presents a practical and efficient solution for T2V generation, combining high performance, scalability, and flexibility for real-world applications. |
|
2025-03-25T00:00:00 | 2503.16924 | Optimized Minimal 3D Gaussian Splatting | [
"Joo Chan Lee",
"Jong Hwan Ko",
"Eunbyung Park"
] | 3D Gaussian Splatting (3DGS) has emerged as a powerful representation for real-time, high-performance rendering, enabling a wide range of applications. However, representing 3D scenes with numerous explicit Gaussian primitives imposes significant storage and memory overhead. Recent studies have shown that high-quality rendering can be achieved with a substantially reduced number of Gaussians when represented with high-precision attributes. Nevertheless, existing 3DGS compression methods still rely on a relatively large number of Gaussians, focusing primarily on attribute compression. This is because a smaller set of Gaussians becomes increasingly sensitive to lossy attribute compression, leading to severe quality degradation. Since the number of Gaussians is directly tied to computational costs, it is essential to reduce the number of Gaussians effectively rather than only optimizing storage. In this paper, we propose Optimized Minimal Gaussians representation (OMG), which significantly reduces storage while using a minimal number of primitives. First, we determine the distinct Gaussian from the near ones, minimizing redundancy without sacrificing quality. Second, we propose a compact and precise attribute representation that efficiently captures both continuity and irregularity among primitives. Additionally, we propose a sub-vector quantization technique for improved irregularity representation, maintaining fast training with a negligible codebook size. Extensive experiments demonstrate that OMG reduces storage requirements by nearly 50% compared to the previous state-of-the-art and enables 600+ FPS rendering while maintaining high rendering quality. Our source code is available at https://maincold2.github.io/omg/. |
|
2025-03-25T00:00:00 | 2503.15879 | Typed-RAG: Type-aware Multi-Aspect Decomposition for Non-Factoid Question Answering | [
"DongGeon Lee",
"Ahjeong Park",
"Hyeri Lee",
"Hyeonseo Nam",
"Yunho Maeng"
] | https://github.com/TeamNLP/Typed-RAG{https: | Non-factoid question-answering (NFQA) poses a significant challenge due to its open-ended nature, diverse intents, and the need for multi-aspect reasoning, which renders conventional factoid QA approaches, including retrieval-augmented generation (RAG), inadequate. Unlike factoid questions, non-factoid questions (NFQs) lack definitive answers and require synthesizing information from multiple sources across various reasoning dimensions. To address these limitations, we introduce Typed-RAG, a type-aware multi-aspect decomposition framework within the RAG paradigm for NFQA. Typed-RAG classifies NFQs into distinct types -- such as debate, experience, and comparison -- and applies aspect-based decomposition to refine retrieval and generation strategies. By decomposing multi-aspect NFQs into single-aspect sub-queries and aggregating the results, Typed-RAG generates more informative and contextually relevant responses. To evaluate Typed-RAG, we introduce Wiki-NFQA, a benchmark dataset covering diverse NFQ types. Experimental results demonstrate that Typed-RAG outperforms baselines, thereby highlighting the importance of type-aware decomposition for effective retrieval and generation in NFQA. Our code and dataset are available at https://github.com/TeamNLP/Typed-RAG{https://github.com/TeamNLP/Typed-RAG}. |
2025-03-25T00:00:00 | 2503.17500 | Variance Control via Weight Rescaling in LLM Pre-training | [
"Louis Owen",
"Abhay Kumar",
"Nilabhra Roy Chowdhury",
"Fabian Güra"
] | https://github.com/bluorion-com/weight_rescaling | The outcome of Large Language Model (LLM) pre-training strongly depends on weight initialization and variance control strategies. Although the importance of initial variance control has been well documented in neural networks in general, the literature on initialization and management of its growth during LLM pre-training, specifically, is somewhat sparse. In this paper, we introduce the Layer Index Rescaling (LIR) weight initialization scheme, and the Target Variance Rescaling (TVR) variance control strategy. Experiments on a 1B parameter LLaMA model demonstrate that better variance management using these techniques yields substantial improvements in downstream task performance (up to 4.6% on common pre-training benchmarks) and reduces extreme activation values, thus mitigating challenges associated with quantization and low-precision training. Our code is available at: https://github.com/bluorion-com/weight_rescaling. |
2025-03-25T00:00:00 | 2503.18102 | AgentRxiv: Towards Collaborative Autonomous Research | [
"Samuel Schmidgall",
"Michael Moor"
] | Progress in scientific discovery is rarely the result of a single "Eureka" moment, but is rather the product of hundreds of scientists incrementally working together toward a common goal. While existing agent workflows are capable of producing research autonomously, they do so in isolation, without the ability to continuously improve upon prior research results. To address these challenges, we introduce AgentRxiv-a framework that lets LLM agent laboratories upload and retrieve reports from a shared preprint server in order to collaborate, share insights, and iteratively build on each other's research. We task agent laboratories to develop new reasoning and prompting techniques and find that agents with access to their prior research achieve higher performance improvements compared to agents operating in isolation (11.4% relative improvement over baseline on MATH-500). We find that the best performing strategy generalizes to benchmarks in other domains (improving on average by 3.3%). Multiple agent laboratories sharing research through AgentRxiv are able to work together towards a common goal, progressing more rapidly than isolated laboratories, achieving higher overall accuracy (13.7% relative improvement over baseline on MATH-500). These findings suggest that autonomous agents may play a role in designing future AI systems alongside humans. We hope that AgentRxiv allows agents to collaborate toward research goals and enables researchers to accelerate discovery. |
|
2025-03-25T00:00:00 | 2503.18813 | Defeating Prompt Injections by Design | [
"Edoardo Debenedetti",
"Ilia Shumailov",
"Tianqi Fan",
"Jamie Hayes",
"Nicholas Carlini",
"Daniel Fabian",
"Christoph Kern",
"Chongyang Shi",
"Andreas Terzis",
"Florian Tramèr"
] | Large Language Models (LLMs) are increasingly deployed in agentic systems that interact with an external environment. However, LLM agents are vulnerable to prompt injection attacks when handling untrusted data. In this paper we propose CaMeL, a robust defense that creates a protective system layer around the LLM, securing it even when underlying models may be susceptible to attacks. To operate, CaMeL explicitly extracts the control and data flows from the (trusted) query; therefore, the untrusted data retrieved by the LLM can never impact the program flow. To further improve security, CaMeL relies on a notion of a capability to prevent the exfiltration of private data over unauthorized data flows. We demonstrate effectiveness of CaMeL by solving 67% of tasks with provable security in AgentDojo [NeurIPS 2024], a recent agentic security benchmark. |
|
2025-03-25T00:00:00 | 2503.18948 | Equivariant Image Modeling | [
"Ruixiao Dong",
"Mengde Xu",
"Zigang Geng",
"Li Li",
"Han Hu",
"Shuyang Gu"
] | https://github.com/drx-code/EquivariantModeling | Current generative models, such as autoregressive and diffusion approaches, decompose high-dimensional data distribution learning into a series of simpler subtasks. However, inherent conflicts arise during the joint optimization of these subtasks, and existing solutions fail to resolve such conflicts without sacrificing efficiency or scalability. We propose a novel equivariant image modeling framework that inherently aligns optimization targets across subtasks by leveraging the translation invariance of natural visual signals. Our method introduces (1) column-wise tokenization which enhances translational symmetry along the horizontal axis, and (2) windowed causal attention which enforces consistent contextual relationships across positions. Evaluated on class-conditioned ImageNet generation at 256x256 resolution, our approach achieves performance comparable to state-of-the-art AR models while using fewer computational resources. Systematic analysis demonstrates that enhanced equivariance reduces inter-task conflicts, significantly improving zero-shot generalization and enabling ultra-long image synthesis. This work establishes the first framework for task-aligned decomposition in generative modeling, offering insights into efficient parameter sharing and conflict-free optimization. The code and models are publicly available at https://github.com/drx-code/EquivariantModeling. |
2025-03-25T00:00:00 | 2503.18878 | I Have Covered All the Bases Here: Interpreting Reasoning Features in Large Language Models via Sparse Autoencoders | [
"Andrey Galichin",
"Alexey Dontsov",
"Polina Druzhinina",
"Anton Razzhigaev",
"Oleg Y. Rogov",
"Elena Tutubalina",
"Ivan Oseledets"
] | https://github.com/AIRI-Institute/SAE-Reasoning | Large Language Models (LLMs) have achieved remarkable success in natural language processing. Recent advances have led to the developing of a new class of reasoning LLMs; for example, open-source DeepSeek-R1 has achieved state-of-the-art performance by integrating deep thinking and complex reasoning. Despite these impressive capabilities, the internal reasoning mechanisms of such models remain unexplored. In this work, we employ Sparse Autoencoders (SAEs), a method to learn a sparse decomposition of latent representations of a neural network into interpretable features, to identify features that drive reasoning in the DeepSeek-R1 series of models. First, we propose an approach to extract candidate ''reasoning features'' from SAE representations. We validate these features through empirical analysis and interpretability methods, demonstrating their direct correlation with the model's reasoning abilities. Crucially, we demonstrate that steering these features systematically enhances reasoning performance, offering the first mechanistic account of reasoning in LLMs. Code available at https://github.com/AIRI-Institute/SAE-Reasoning |
2025-03-25T00:00:00 | 2503.18886 | CFG-Zero*: Improved Classifier-Free Guidance for Flow Matching Models | [
"Weichen Fan",
"Amber Yijia Zheng",
"Raymond A. Yeh",
"Ziwei Liu"
] | Classifier-Free Guidance (CFG) is a widely adopted technique in diffusion/flow models to improve image fidelity and controllability. In this work, we first analytically study the effect of CFG on flow matching models trained on Gaussian mixtures where the ground-truth flow can be derived. We observe that in the early stages of training, when the flow estimation is inaccurate, CFG directs samples toward incorrect trajectories. Building on this observation, we propose CFG-Zero*, an improved CFG with two contributions: (a) optimized scale, where a scalar is optimized to correct for the inaccuracies in the estimated velocity, hence the * in the name; and (b) zero-init, which involves zeroing out the first few steps of the ODE solver. Experiments on both text-to-image (Lumina-Next, Stable Diffusion 3, and Flux) and text-to-video (Wan-2.1) generation demonstrate that CFG-Zero* consistently outperforms CFG, highlighting its effectiveness in guiding Flow Matching models. (Code is available at github.com/WeichenFan/CFG-Zero-star) |
|
2025-03-25T00:00:00 | 2503.18470 | MetaSpatial: Reinforcing 3D Spatial Reasoning in VLMs for the Metaverse | [
"Zhenyu Pan",
"Han Liu"
] | https://github.com/PzySeere/MetaSpatial | We present MetaSpatial, the first reinforcement learning (RL)-based framework designed to enhance 3D spatial reasoning in vision-language models (VLMs), enabling real-time 3D scene generation without the need for hard-coded optimizations. MetaSpatial addresses two core challenges: (i) the lack of internalized 3D spatial reasoning in VLMs, which limits their ability to generate realistic layouts, and (ii) the inefficiency of traditional supervised fine-tuning (SFT) for layout generation tasks, as perfect ground truth annotations are unavailable. Our key innovation is a multi-turn RL-based optimization mechanism that integrates physics-aware constraints and rendered image evaluations, ensuring generated 3D layouts are coherent, physically plausible, and aesthetically consistent. Methodologically, MetaSpatial introduces an adaptive, iterative reasoning process, where the VLM refines spatial arrangements over multiple turns by analyzing rendered outputs, improving scene coherence progressively. Empirical evaluations demonstrate that MetaSpatial significantly enhances the spatial consistency and formatting stability of various scale models. Post-training, object placements are more realistic, aligned, and functionally coherent, validating the effectiveness of RL for 3D spatial reasoning in metaverse, AR/VR, digital twins, and game development applications. Our code, data, and training pipeline are publicly available at https://github.com/PzySeere/MetaSpatial. |
2025-03-25T00:00:00 | 2503.18494 | Verbal Process Supervision Elicits Better Coding Agents | [
"Hao-Yuan Chen",
"Cheng-Pong Huang",
"Jui-Ming Yao"
] | The emergence of large language models and their applications as AI agents have significantly advanced state-of-the-art code generation benchmarks, transforming modern software engineering tasks. However, even with test-time computed reasoning models, these systems still struggle with complex software engineering challenges. This work introduces CURA, a code understanding and reasoning agent system enhanced with verbal process supervision (VPS), achieving a 3.65\% improvement over baseline models on challenging benchmarks like BigCodeBench. Furthermore, CURA, when paired with the o3-mini model and VPS techniques, attains state-of-the-art performance. This work represents a step forward in integrating reasoning-driven architectures with LLM-based code generation, enabling agentic reasoning for language models to solve complex software engineering tasks. |
|
2025-03-25T00:00:00 | 2503.18033 | OmnimatteZero: Training-free Real-time Omnimatte with Pre-trained Video Diffusion Models | [
"Dvir Samuel",
"Matan Levy",
"Nir Darshan",
"Gal Chechik",
"Rami Ben-Ari"
] | Omnimatte aims to decompose a given video into semantically meaningful layers, including the background and individual objects along with their associated effects, such as shadows and reflections. Existing methods often require extensive training or costly self-supervised optimization. In this paper, we present OmnimatteZero, a training-free approach that leverages off-the-shelf pre-trained video diffusion models for omnimatte. It can remove objects from videos, extract individual object layers along with their effects, and composite those objects onto new videos. We accomplish this by adapting zero-shot image inpainting techniques for video object removal, a task they fail to handle effectively out-of-the-box. We then show that self-attention maps capture information about the object and its footprints and use them to inpaint the object's effects, leaving a clean background. Additionally, through simple latent arithmetic, object layers can be isolated and recombined seamlessly with new video layers to produce new videos. Evaluations show that OmnimatteZero not only achieves superior performance in terms of background reconstruction but also sets a new record for the fastest Omnimatte approach, achieving real-time performance with minimal frame runtime. |
|
2025-03-25T00:00:00 | 2503.18352 | Diffusion-4K: Ultra-High-Resolution Image Synthesis with Latent Diffusion Models | [
"Jinjin Zhang",
"Qiuyu Huang",
"Junjie Liu",
"Xiefan Guo",
"Di Huang"
] | In this paper, we present Diffusion-4K, a novel framework for direct ultra-high-resolution image synthesis using text-to-image diffusion models. The core advancements include: (1) Aesthetic-4K Benchmark: addressing the absence of a publicly available 4K image synthesis dataset, we construct Aesthetic-4K, a comprehensive benchmark for ultra-high-resolution image generation. We curated a high-quality 4K dataset with carefully selected images and captions generated by GPT-4o. Additionally, we introduce GLCM Score and Compression Ratio metrics to evaluate fine details, combined with holistic measures such as FID, Aesthetics and CLIPScore for a comprehensive assessment of ultra-high-resolution images. (2) Wavelet-based Fine-tuning: we propose a wavelet-based fine-tuning approach for direct training with photorealistic 4K images, applicable to various latent diffusion models, demonstrating its effectiveness in synthesizing highly detailed 4K images. Consequently, Diffusion-4K achieves impressive performance in high-quality image synthesis and text prompt adherence, especially when powered by modern large-scale diffusion models (e.g., SD3-2B and Flux-12B). Extensive experimental results from our benchmark demonstrate the superiority of Diffusion-4K in ultra-high-resolution image synthesis. |
|
2025-03-25T00:00:00 | 2503.18018 | Lost in Cultural Translation: Do LLMs Struggle with Math Across Cultural Contexts? | [
"Aabid Karim",
"Abdul Karim",
"Bhoomika Lohana",
"Matt Keon",
"Jaswinder Singh",
"Abdul Sattar"
] | https://github.com/akarim23131/Lost_in_Cultural_Translation | Large Language Models (LLMs) have significantly advanced various fields, particularly coding, mathematical reasoning, and logical problem solving. However, a critical question remains: Do these mathematical reasoning abilities persist when LLMs are presented with culturally adapted math problems? Specifically, how do LLMs perform when faced with math problems embedded in cultural contexts that have no significant representation in main stream web-scale AI training data? To explore this, we generated six synthetic cultural datasets from GSM8K, a widely used benchmark for assessing LLMs' mathematical reasoning skills. While preserving the mathematical logic and numerical values of the original GSM8K test set, we modify cultural elements such as personal names, food items, place names, etc. These culturally adapted datasets provide a more reliable framework for evaluating LLMs' mathematical reasoning under shifting cultural contexts. Our findings reveal that LLMs struggle with math problems when cultural references change, even though the underlying mathematical structure remains constant. Smaller models exhibit greater performance drops compared to larger models. Interestingly, our results also suggest that cultural familiarity can enhance mathematical reasoning. Even models with no explicit mathematical training but exposure to relevant cultural contexts sometimes outperform larger, mathematically proficient models on culturally embedded math problems. This study highlights the impact of cultural context on the mathematical reasoning abilities of LLMs, underscoring the need for more diverse and representative training data to improve robustness in real-world applications. The benchmark data sets and script for reproducing the results are available at https://github.com/akarim23131/Lost_in_Cultural_Translation |
2025-03-25T00:00:00 | 2503.16426 | DynamicVis: An Efficient and General Visual Foundation Model for Remote Sensing Image Understanding | [
"Keyan Chen",
"Chenyang Liu",
"Bowen Chen",
"Wenyuan Li",
"Zhengxia Zou",
"Zhenwei Shi"
] | The advancement of remote sensing technology has improved the spatial resolution of satellite imagery, facilitating more detailed visual representations for diverse interpretations. However, existing methods exhibit limited generalization capabilities across varied applications. While some contemporary foundation models demonstrate potential, they are hindered by insufficient cross-task adaptability and primarily process low-resolution imagery of restricted sizes, thus failing to fully exploit high-resolution data or leverage comprehensive large-scene semantics. Crucially, remote sensing imagery differs fundamentally from natural images, as key foreground targets (eg., maritime objects, artificial structures) often occupy minimal spatial proportions (~1%) and exhibit sparse distributions. Efficiently modeling cross-task generalizable knowledge from lengthy 2D tokens (~100,000) poses a significant challenge yet remains critical for remote sensing image understanding. Motivated by the selective attention mechanisms inherent to the human visual system, we propose DynamicVis, a dynamic visual perception foundation model for remote sensing imagery. The framework integrates a novel dynamic region perception backbone based on the selective state space model, which strategically balances localized detail extraction with global contextual integration, enabling computationally efficient encoding of large-scale data while maintaining architectural scalability. To enhance cross-task knowledge transferring, we introduce a multi-instance learning paradigm utilizing meta-embedding representations, trained on million-scale region-level annotations. Evaluations across nine downstream tasks demonstrate the model's versatility. DynamicVis achieves multi-level feature modeling with exceptional efficiency, processing (2048x2048) pixels with 97 ms latency (6% of ViT's) and 833 MB GPU memory (3% of ViT's). |
|
2025-03-25T00:00:00 | 2503.17811 | Feather-SQL: A Lightweight NL2SQL Framework with Dual-Model Collaboration Paradigm for Small Language Models | [
"Wenqi Pei",
"Hailing Xu",
"Hengyuan Zhao",
"Shizheng Hou",
"Han Chen",
"Zining Zhang",
"Pingyi Luo",
"Bingsheng He"
] | Natural Language to SQL (NL2SQL) has seen significant advancements with large language models (LLMs). However, these models often depend on closed-source systems and high computational resources, posing challenges in data privacy and deployment. In contrast, small language models (SLMs) struggle with NL2SQL tasks, exhibiting poor performance and incompatibility with existing frameworks. To address these issues, we introduce Feather-SQL, a new lightweight framework tailored for SLMs. Feather-SQL improves SQL executability and accuracy through 1) schema pruning and linking, 2) multi-path and multi-candidate generation. Additionally, we introduce the 1+1 Model Collaboration Paradigm, which pairs a strong general-purpose chat model with a fine-tuned SQL specialist, combining strong analytical reasoning with high-precision SQL generation. Experimental results on BIRD demonstrate that Feather-SQL improves NL2SQL performance on SLMs, with around 10% boost for models without fine-tuning. The proposed paradigm raises the accuracy ceiling of SLMs to 54.76%, highlighting its effectiveness. |
|
2025-03-25T00:00:00 | 2503.17760 | CODA: Repurposing Continuous VAEs for Discrete Tokenization | [
"Zeyu Liu",
"Zanlin Ni",
"Yeguo Hua",
"Xin Deng",
"Xiao Ma",
"Cheng Zhong",
"Gao Huang"
] | Discrete visual tokenizers transform images into a sequence of tokens, enabling token-based visual generation akin to language models. However, this process is inherently challenging, as it requires both compressing visual signals into a compact representation and discretizing them into a fixed set of codes. Traditional discrete tokenizers typically learn the two tasks jointly, often leading to unstable training, low codebook utilization, and limited reconstruction quality. In this paper, we introduce CODA(COntinuous-to-Discrete Adaptation), a framework that decouples compression and discretization. Instead of training discrete tokenizers from scratch, CODA adapts off-the-shelf continuous VAEs -- already optimized for perceptual compression -- into discrete tokenizers via a carefully designed discretization process. By primarily focusing on discretization, CODA ensures stable and efficient training while retaining the strong visual fidelity of continuous VAEs. Empirically, with 6 times less training budget than standard VQGAN, our approach achieves a remarkable codebook utilization of 100% and notable reconstruction FID (rFID) of 0.43 and 1.34 for 8 times and 16 times compression on ImageNet 256times 256 benchmark. |
|
2025-03-25T00:00:00 | 2503.18071 | Mind with Eyes: from Language Reasoning to Multimodal Reasoning | [
"Zhiyu Lin",
"Yifei Gao",
"Xian Zhao",
"Yunfan Yang",
"Jitao Sang"
] | Language models have recently advanced into the realm of reasoning, yet it is through multimodal reasoning that we can fully unlock the potential to achieve more comprehensive, human-like cognitive capabilities. This survey provides a systematic overview of the recent multimodal reasoning approaches, categorizing them into two levels: language-centric multimodal reasoning and collaborative multimodal reasoning. The former encompasses one-pass visual perception and active visual perception, where vision primarily serves a supporting role in language reasoning. The latter involves action generation and state update within reasoning process, enabling a more dynamic interaction between modalities. Furthermore, we analyze the technical evolution of these methods, discuss their inherent challenges, and introduce key benchmark tasks and evaluation metrics for assessing multimodal reasoning performance. Finally, we provide insights into future research directions from the following two perspectives: (i) from visual-language reasoning to omnimodal reasoning and (ii) from multimodal reasoning to multimodal agents. This survey aims to provide a structured overview that will inspire further advancements in multimodal reasoning research. |
|
2025-03-25T00:00:00 | 2503.14774 | Revisiting Image Fusion for Multi-Illuminant White-Balance Correction | [
"David Serrano-Lozano",
"Aditya Arora",
"Luis Herranz",
"Konstantinos G. Derpanis",
"Michael S. Brown",
"Javier Vazquez-Corral"
] | White balance (WB) correction in scenes with multiple illuminants remains a persistent challenge in computer vision. Recent methods explored fusion-based approaches, where a neural network linearly blends multiple sRGB versions of an input image, each processed with predefined WB presets. However, we demonstrate that these methods are suboptimal for common multi-illuminant scenarios. Additionally, existing fusion-based methods rely on sRGB WB datasets lacking dedicated multi-illuminant images, limiting both training and evaluation. To address these challenges, we introduce two key contributions. First, we propose an efficient transformer-based model that effectively captures spatial dependencies across sRGB WB presets, substantially improving upon linear fusion techniques. Second, we introduce a large-scale multi-illuminant dataset comprising over 16,000 sRGB images rendered with five different WB settings, along with WB-corrected images. Our method achieves up to 100\% improvement over existing techniques on our new multi-illuminant image fusion dataset. |
|
2025-03-25T00:00:00 | 2503.13074 | Rethinking Image Evaluation in Super-Resolution | [
"Shaolin Su",
"Josep M. Rocafort",
"Danna Xue",
"David Serrano-Lozano",
"Lei Sun",
"Javier Vazquez-Corral"
] | While recent advancing image super-resolution (SR) techniques are continually improving the perceptual quality of their outputs, they can usually fail in quantitative evaluations. This inconsistency leads to a growing distrust in existing image metrics for SR evaluations. Though image evaluation depends on both the metric and the reference ground truth (GT), researchers typically do not inspect the role of GTs, as they are generally accepted as `perfect' references. However, due to the data being collected in the early years and the ignorance of controlling other types of distortions, we point out that GTs in existing SR datasets can exhibit relatively poor quality, which leads to biased evaluations. Following this observation, in this paper, we are interested in the following questions: Are GT images in existing SR datasets 100% trustworthy for model evaluations? How does GT quality affect this evaluation? And how to make fair evaluations if there exist imperfect GTs? To answer these questions, this paper presents two main contributions. First, by systematically analyzing seven state-of-the-art SR models across three real-world SR datasets, we show that SR performances can be consistently affected across models by low-quality GTs, and models can perform quite differently when GT quality is controlled. Second, we propose a novel perceptual quality metric, Relative Quality Index (RQI), that measures the relative quality discrepancy of image pairs, thus issuing the biased evaluations caused by unreliable GTs. Our proposed model achieves significantly better consistency with human opinions. We expect our work to provide insights for the SR community on how future datasets, models, and metrics should be developed. |
|
2025-03-25T00:00:00 | 2503.18674 | Human Motion Unlearning | [
"Edoardo De Matteis",
"Matteo Migliarini",
"Alessio Sampieri",
"Indro Spinelli",
"Fabio Galasso"
] | We introduce the task of human motion unlearning to prevent the synthesis of toxic animations while preserving the general text-to-motion generative performance. Unlearning toxic motions is challenging as those can be generated from explicit text prompts and from implicit toxic combinations of safe motions (e.g., ``kicking" is ``loading and swinging a leg"). We propose the first motion unlearning benchmark by filtering toxic motions from the large and recent text-to-motion datasets of HumanML3D and Motion-X. We propose baselines, by adapting state-of-the-art image unlearning techniques to process spatio-temporal signals. Finally, we propose a novel motion unlearning model based on Latent Code Replacement, which we dub LCR. LCR is training-free and suitable to the discrete latent spaces of state-of-the-art text-to-motion diffusion models. LCR is simple and consistently outperforms baselines qualitatively and quantitatively. Project page: https://www.pinlab.org/hmu{https://www.pinlab.org/hmu}. |
|
2025-03-25T00:00:00 | 2503.16709 | QuartDepth: Post-Training Quantization for Real-Time Depth Estimation on the Edge | [
"Xuan Shen",
"Weize Ma",
"Jing Liu",
"Changdi Yang",
"Rui Ding",
"Quanyi Wang",
"Henghui Ding",
"Wei Niu",
"Yanzhi Wang",
"Pu Zhao",
"Jun Lin",
"Jiuxiang Gu"
] | https://github.com/shawnricecake/quart-depth | Monocular Depth Estimation (MDE) has emerged as a pivotal task in computer vision, supporting numerous real-world applications. However, deploying accurate depth estimation models on resource-limited edge devices, especially Application-Specific Integrated Circuits (ASICs), is challenging due to the high computational and memory demands. Recent advancements in foundational depth estimation deliver impressive results but further amplify the difficulty of deployment on ASICs. To address this, we propose QuartDepth which adopts post-training quantization to quantize MDE models with hardware accelerations for ASICs. Our approach involves quantizing both weights and activations to 4-bit precision, reducing the model size and computation cost. To mitigate the performance degradation, we introduce activation polishing and compensation algorithm applied before and after activation quantization, as well as a weight reconstruction method for minimizing errors in weight quantization. Furthermore, we design a flexible and programmable hardware accelerator by supporting kernel fusion and customized instruction programmability, enhancing throughput and efficiency. Experimental results demonstrate that our framework achieves competitive accuracy while enabling fast inference and higher energy efficiency on ASICs, bridging the gap between high-performance depth estimation and practical edge-device applicability. Code: https://github.com/shawnricecake/quart-depth |
2025-03-25T00:00:00 | 2503.18406 | Instruct-CLIP: Improving Instruction-Guided Image Editing with Automated Data Refinement Using Contrastive Learning | [
"Sherry X. Chen",
"Misha Sra",
"Pradeep Sen"
] | https://github.com/SherryXTChen/Instruct-CLIP.git | Although natural language instructions offer an intuitive way to guide automated image editing, deep-learning models often struggle to achieve high-quality results, largely due to challenges in creating large, high-quality training datasets. Previous work has typically relied on text-toimage (T2I) generative models to produce pairs of original and edited images that simulate the input/output of an instruction-guided image-editing model. However, these image pairs often fail to align with the specified edit instructions due to the limitations of T2I models, which negatively impacts models trained on such datasets. To address this, we present Instruct-CLIP, a self-supervised method that learns the semantic changes between original and edited images to refine and better align the instructions in existing datasets. Furthermore, we adapt Instruct-CLIP to handle noisy latent images and diffusion timesteps so that it can be used to train latent diffusion models (LDMs) [19] and efficiently enforce alignment between the edit instruction and the image changes in latent space at any step of the diffusion pipeline. We use Instruct-CLIP to correct the InstructPix2Pix dataset and get over 120K refined samples we then use to fine-tune their model, guided by our novel Instruct-CLIP-based loss function. The resulting model can produce edits that are more aligned with the given instructions. Our code and dataset are available at https://github.com/SherryXTChen/Instruct-CLIP.git. |
Subsets and Splits