AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

MARS: Multimodal Adaptive Reasoning Model for Avoiding Overthinking

Tan Yue, Qiong Wu, Dongyan Zhao

AAAI 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Chain-of-thought multimodal reasoning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes MARS, a multimodal LLM reasoning model that adaptively adjusts its reasoning strategy to question difficulty to avoid overthinking and reduce computational cost.

Problem

Multimodal large language models often produce excessively long reasoning chains, causing high computational cost and inefficiency on vision-language tasks.

Approach

Enables adaptive adjustment of the reasoning strategy based on question difficulty so that easy questions receive shorter reasoning while harder ones get fuller reasoning; the specific mechanism is not stated in the abstract provided.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add CoT Masking Learning (stage 1) (ablation)MARS (Ins.+RL) · SciChart (out-of-domain)R-Acc: 38.23 -> 41.68Improved
add IB-GRPO in place of raw GRPO (ablation)MARS three-stage · SciChart (out-of-domain)average tokens: 256.68 -> 95.81Improved
add IB-GRPO in place of raw GRPO (ablation)MARS three-stage · SciChart (out-of-domain)R-Acc: 41.68 -> 42.31Improved
add RL stage (with IB-GRPO) after Mask+Ins. (ablation)MARS · SciChart (out-of-domain)average tokens: 218.65 -> 95.81Improved
add RL stage (with IB-GRPO) after Mask+Ins. (ablation)MARS · SciChart (out-of-domain)R-Acc: 40.68 -> 42.31Improved
add Adaptive Reasoning Instruction Learning (ablation)MARS (Mask+RL) · SciChart (out-of-domain)average tokens: 686.17 -> 95.81Improved
add Adaptive Reasoning Instruction Learning (ablation)MARS (Mask+RL) · SciChart (out-of-domain)R-Acc: 41.15 -> 42.31Improved
apply MARS training to Mimo-VL-7B-RLMimo-VL-7B-RL · SciChart (out-of-domain)average tokens: 980.09 -> 95.81Improved
apply MARS training to Mimo-VL-7B-RLMimo-VL-7B-RL · SciChart (out-of-domain)R-Acc: 41.77 -> 42.31Improved
apply MARS on FWHM questionsMimo-VL-7B-RL · SciChart (out-of-domain)R-Acc: 16.95 -> 15.39Hurt
apply MARS on Peak Position questionsMimo-VL-7B-RL · SciChart (out-of-domain)R-Acc: 48.99 -> 48.56No clear effect
thinking mode vs non-thinkingKimi-VL-A3B · SciChart (out-of-domain)R-Acc: 25.10 -> 27.78Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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