{"slug": "videokr-towards-knowledge-and-reasoning-intensive-video-understanding", "title": "VideoKR: Towards Knowledge- and Reasoning-Intensive Video Understanding", "summary": "Researchers introduced VideoKR, a large-scale training corpus of 315,000 video reasoning examples over 145,000 expert-domain videos designed to strengthen knowledge- and reasoning-intensive video understanding. The team developed a human-in-the-loop generation pipeline and a new expert-annotated benchmark, VideoKR-Eval, to ensure genuine video comprehension rather than reliance on textual shortcuts. Experiments showed that models post-trained on VideoKR outperformed prior approaches on knowledge-intensive video reasoning while remaining competitive on general video reasoning, demonstrating that data design is a key driver of progress in the field.", "body_md": "arXiv:2606.05259v1 Announce Type: new\nAbstract: We introduce VideoKR, the first large-scale training corpus specifically designed to strengthen knowledge- and reasoning-intensive video understanding. It comprises 315K video reasoning examples over 145K newly collected, CC-licensed, expert-domain videos. We develop a human-in-the-loop, skill-oriented example generation pipeline that targets progressively deeper video reasoning capabilities while ensuring the difficulty, diversity, and reliability of both the examples and their CoT rationales. We also curate VideoKR-Eval, a new expert-annotated benchmark where questions require genuine video understanding and knowledge-intensive reasoning rather than textual shortcuts. Our experiments show that, under a standard SFT$\\rightarrow$GRPO pipeline, models post-trained on VideoKR outperform prior post-training approaches on knowledge-intensive video reasoning while remaining competitive on general video reasoning, highlighting data design as a key driver of progress in video reasoning. We further conduct comprehensive ablations to isolate the contributions of VideoKR, providing actionable insights for future work.", "url": "https://wpnews.pro/news/videokr-towards-knowledge-and-reasoning-intensive-video-understanding", "canonical_source": "https://arxiv.org/abs/2606.05259", "published_at": "2026-06-05 04:00:00+00:00", "updated_at": "2026-06-05 04:16:58.211460+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "computer-vision", "large-language-models", "ai-research"], "entities": ["VideoKR", "VideoKR-Eval", "SFT", "GRPO", "CC"], "alternates": {"html": "https://wpnews.pro/news/videokr-towards-knowledge-and-reasoning-intensive-video-understanding", "markdown": "https://wpnews.pro/news/videokr-towards-knowledge-and-reasoning-intensive-video-understanding.md", "text": "https://wpnews.pro/news/videokr-towards-knowledge-and-reasoning-intensive-video-understanding.txt", "jsonld": "https://wpnews.pro/news/videokr-towards-knowledge-and-reasoning-intensive-video-understanding.jsonld"}}