Seeing Clearly, Answering Incorrectly: A Multimodal Robustness Benchmark for Evaluating MLLMs on Leading Questions
Paper
•
2406.10638
•
Published
Zhengyang Liang
chr1ce
·
AI & ML interests
None yet
Recent Activity
upvoted
a
paper
2 days ago
Video-BrowseComp: Benchmarking Agentic Video Research on Open Web
updated
a Space
about 1 month ago
UniVA-Agent/README
upvoted
a
paper
about 1 month ago
General Agentic Memory Via Deep Research
Organizations
authored
6
papers
about 2 months ago
MomentSeeker: A Comprehensive Benchmark and A Strong Baseline For Moment Retrieval Within Long Videos
Paper
•
2502.12558
•
Published
Any Information Is Just Worth One Single Screenshot: Unifying Search With Visualized Information Retrieval
Paper
•
2502.11431
•
Published
Video-XL-2: Towards Very Long-Video Understanding Through Task-Aware KV Sparsification
Paper
•
2506.19225
•
Published
TimeScope: Towards Task-Oriented Temporal Grounding In Long Videos
Paper
•
2509.26360
•
Published
UniVA: Universal Video Agent towards Open-Source Next-Generation Video Generalist
Paper
•
2511.08521
•
Published
•
37