Resources & Downloads
Open-source codebases, datasets and benchmarks we've released, survey write-ups, LaTeX templates I've built for my own papers and posters, and featured talk videos.
Featured Videos
Open-source codebases, datasets and benchmarks we've released, survey write-ups, LaTeX templates I've built for my own papers and posters, and featured talk videos.
140 tasks across 12 domains testing whether AI agents can invent generalizable, scalable ML methods rather than only apply existing ones.
A benchmark for AI agents on ~100 terminal tasks.
A benchmark for visual reasoning that evaluates fundamental visual skills independent of language shortcuts.
A 1,000-example benchmark for evaluating models that generate both images and text with rubric-based scoring.
An open-ended benchmark for challenging computer science problems with objective, fine-grained evaluation.
First benchmark dedicated to Dense Video Understanding, focusing on QA-driven high-frame-rate comprehension.
Models like o3-high, o4-mini, and Gemini 2.5 Pro score 0% on hard competitive programming problems.
A massive benchmark designed to evaluate LMMs in understanding Multi-Discipline Lectures.
1M reasoning samples about causal event relationships with fine-grained, timestamped descriptions of untrimmed videos.
First benchmark for reasoning-informed visual editing across four reasoning types: temporal, causal, spatial, logical.
Over 20k image pairs for training a language-guided reward model for text-to-image alignment with scientific knowledge.
First benchmark for detailed video captioning — 1k+ videos with significantly longer captions plus training recipes.
Human pose estimation dataset with calibrated radar ADC data, 4D radar tensors, stereo RGB images, and LiDAR.
Manually labeled long-video QA and caption dataset — 1,000 videos, each longer than ten thousand frames.
A benchmark for AI agents on ~100 terminal tasks.
Zero-shot visual object tracking with motion-aware memory, built on SAM.
Evaluation suite accelerating the development of large multimodal models.
Large multimodal models for long-form video understanding with memory mechanism.
Text-driven, consistency-aware diffusion video editing.