Zebra-CoT: A Dataset for Interleaved Vision Language Reasoning Paper • 2507.16746 • Published about 16 hours ago • 7
PHTest Collection Automatic Pseudo-Harmful Prompt Generation for Evaluating False Refusals in Large Language Models • 3 items • Updated Sep 24, 2024 • 1
Easy2Hard-Bench Collection Easy2Hard-Bench offers six datasets with continuous difficulty ratings, enabling profiling of LLM performance and generalization across difficulties. • 7 items • Updated Jul 3, 2024 • 1
TraceVLA Collection TraceVLA: Visual Trace Prompting Enhances Spatial-Temporal Awareness for Generalist Robotic Policies • 4 items • Updated Jan 7 • 3
WAVES Collection Benchmarking the Robustness of Image Watermarks. Under development. Data will be released soon. • 2 items • Updated Jan 24, 2024 • 3
Recurrent Models Collection These are checkpoints for recurrent LLMs developed to scale test-time compute by recurring in latent space. • 15 items • Updated May 21 • 9
Cross-Modal Safety Alignment: Is textual unlearning all you need? Paper • 2406.02575 • Published May 27, 2024 • 1
Model Tampering Attacks Enable More Rigorous Evaluations of LLM Capabilities Paper • 2502.05209 • Published Feb 3 • 1
MORSE-500: A Programmatically Controllable Video Benchmark to Stress-Test Multimodal Reasoning Paper • 2506.05523 • Published Jun 5 • 34
Cross-Modal Safety Alignment: Is textual unlearning all you need? Paper • 2406.02575 • Published May 27, 2024 • 1
Model Tampering Attacks Enable More Rigorous Evaluations of LLM Capabilities Paper • 2502.05209 • Published Feb 3 • 1