AI Paper Atlas

Research map / Video / Video-language understanding

CausalStep: A Benchmark for Explicit Stepwise Causal Reasoning in Videos

Xuchen Li, Xuzhao Li, Shiyu Hu, Kaiqi Huang, Wentao Zhang

AAAI 2026 · Video › Video-language understanding · approach: Long video understanding · 10 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
human participants vs best model (o4-mini) on chain successo4-mini · CausalStepCSR (%): 51 (o4-mini) vs 79 (human)Hurt
isolated causal question (ICRA) vs descriptive question (DUA) for best modelo4-mini · CausalStepaccuracy: DUA 85.2 -> ICRA 39.8Hurt
proprietary instead of open-source modelsMLLMs · CausalStepCSR (%): 29 (Gemma3-27b) -> 51 (o4-mini)Improved
scale Qwen2.5-VL from 7B to 72BQwen2.5-VL · CausalStepCSR (%): 16 -> 26Improved
scale InternVL3 from 8B to 38BInternVL3 · CausalStepCSR (%): 19 -> 24Improved
scale Gemma3 from 12B to 27BGemma3 · CausalStepCSR (%): 21 -> 29Improved
add reasoning/thinking mode to Gemini-2.0-FlashGemini-2.0-Flash · CausalStepCSR (%): 31 -> 41Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy