Pratyay Banerjee

Neilblaze's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

upvoted 5 papers 1 day ago

CommVQ: Commutative Vector Quantization for KV Cache Compression

Paper • 2506.18879 • Published 4 days ago • 5

Enhancing Step-by-Step and Verifiable Medical Reasoning in MLLMs

Paper • 2506.16962 • Published 8 days ago • 9

SlimMoE: Structured Compression of Large MoE Models via Expert Slimming and Distillation

Paper • 2506.18349 • Published 5 days ago • 9

OAgents: An Empirical Study of Building Effective Agents

Paper • 2506.15741 • Published 10 days ago • 31

Chain-of-Experts: Unlocking the Communication Power of Mixture-of-Experts Models

Paper • 2506.18945 • Published 5 days ago • 33

upvoted 15 papers 4 days ago

Inherently Faithful Attention Maps for Vision Transformers

Paper • 2506.08915 • Published 17 days ago • 4

AbstentionBench: Reasoning LLMs Fail on Unanswerable Questions

Paper • 2506.09038 • Published 17 days ago • 7

Med-PRM: Medical Reasoning Models with Stepwise, Guideline-verified Process Rewards

Paper • 2506.11474 • Published 15 days ago • 16

Feedback Friction: LLMs Struggle to Fully Incorporate External Feedback

Paper • 2506.11930 • Published 14 days ago • 53

AI Agent Behavioral Science

Paper • 2506.06366 • Published 24 days ago • 10

Language Surgery in Multilingual Large Language Models

Paper • 2506.12450 • Published 14 days ago • 16

AR-RAG: Autoregressive Retrieval Augmentation for Image Generation

Paper • 2506.06962 • Published 20 days ago • 28

Ego-R1: Chain-of-Tool-Thought for Ultra-Long Egocentric Video Reasoning

Paper • 2506.13654 • Published 11 days ago • 42

Wait, We Don't Need to "Wait"! Removing Thinking Tokens Improves Reasoning Efficiency

Paper • 2506.08343 • Published 18 days ago • 46

DoTA-RAG: Dynamic of Thought Aggregation RAG

Paper • 2506.12571 • Published 13 days ago • 47

DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents

Paper • 2506.11763 • Published 15 days ago • 58

MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention

Paper • 2506.13585 • Published 11 days ago • 240

Mixture-of-Experts Meets In-Context Reinforcement Learning

Paper • 2506.05426 • Published 23 days ago • 5

Universal Jailbreak Suffixes Are Strong Attention Hijackers

Paper • 2506.12880 • Published 12 days ago • 5

AgentSynth: Scalable Task Generation for Generalist Computer-Use Agents

Paper • 2506.14205 • Published 11 days ago • 6