AI Paper Atlas

Research map / Video / Video-language understanding

MMDuet2: Enhancing Proactive Interaction of Video MLLMs with Multi-Turn Reinforcement Learning

Yueqian Wang, Songxiang Liu, Disong Wang, Nuo Xu, Wan Guanglu, Huishuai Zhang, Dongyan Zhao

ICLR 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 26 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
add multi-turn GRPO RL with PAUC-based reward after SFTMMDuet2 SFT · ProactiveVideoQA [WEB]PAUC: 37.6 -> 53.3Improved
add RL after SFTMMDuet2 SFT · ProactiveVideoQA [EGO]PAUC: 26.4 -> 33.6Improved
add RL after SFTMMDuet2 SFT · StreamingBench Proactive Outputaccuracy: 19.59 -> 34.69Improved
add replication reward r_rep (ablation)MMDuet2 · ProactiveVideoQA [WEB]duplicate proportion: 17.3 -> 4.2Improved
add in-span reward r_in_span (ablation)MMDuet2 · ProactiveVideoQA [EGO]feasibilityImproved
add prefix reward r_pfx (ablation)MMDuet2 · ProactiveVideoQA [EGO]PAUC: 27.5 -> 33.6Improved
add r_pfx (ablation)MMDuet2 · ProactiveVideoQA [WEB]PAUC: 53.0 -> 53.3No clear effect
reduce inference frame interval from 2 s to 1 s (ablation)MMDuet2 RL · ProactiveVideoQA [WEB]PAUC: 44.2 -> 53.3Improved
SFT frame interval 1 s instead of 2 s (ablation)MMDuet2 SFT · ProactiveVideoQAbehaviorHurt
RL frame interval 1 s vs 2 s (ablation)MMDuet2 RL · ProactiveVideoQA [EGO]PAUC: 33.6 -> 34.7No clear effect
proactive SFT+RL trainingQwen2.5-VL 3B · Video-MME (w/o sub)accuracy: 57.3 -> 58.1No clear effect
proactive SFT+RL trainingQwen2.5-VL 3B · LongVideoBenchaccuracy: 53.1 -> 52.7No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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