arXiv Computer Vision October 1, 2026 By Pengzhan Sun, Shiu-hong Kao, Shijie Li, Yongyi Su, Junbin Xiao, Arjun Reddy Akula, Angela Yao

Aligning Thoughts with Answers: Probability Rewards to Tame Thinking Drift

Read the original on arXiv Computer Vision →

The paper introduces Rita, a reinforcement learning framework that addresses thinking drift in vision‑language models by enforcing consistency between reasoning and answers. Rita employs two reasoning‑label‑free rewards—thinking and consistency rewards—derived from the conditional probability of reference answers, and uses a difficulty‑aware data filtering strategy to select informative samples for training. Experiments on EgoIntention and RefEgo‑Int benchmarks demonstrate that Rita outperforms both supervised fine‑tuning and vanilla RL‑fine‑tuned approaches.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computer Vision.

Hugging Face Trending Papers

4d ago

Aligning Thoughts with Answers: Probability Rewards to Tame Thinking Drift

This paper studies \textbf{thinking--answer consistency} in vision-language models. We focus on Visual Intention Grounding, where a model infers a target object based on a human intention query and pr...

llms reinforcement-learning fine-tuning multimodal benchmarks

arXiv AI

2d ago

On Language Drift during RLVR Post-Training

arXiv:2610.02015v1 Announce Type: cross Abstract: Recent advances in LLM reasoning models---driven primarily by the paradigm of post-training via reinforcement learning with verifiable reward (RLVR)-...

By Michael Sullivan, Alexander Koller

llms reinforcement-learning fine-tuning

arXiv AI

Jun 30

Be Faithful When Response: Returning Fluent and Grounded Answers for Vision-Language Models Reinforcement Learning

arXiv:2606. 29984v1 Announce Type: new Abstract: Reinforcement Learning (RL) is an important paradigm for improving the reasoning capabilities of Vision-Language Models (VLMs).

By Peng, Lee, Yin Zhang, Yanglin Zhang, Haonan Wu, Zishan Liu, Ruoxi Zang, Xin Zhu, Jiayin Zheng, Jian Yao, Zefeng Ji, Fei Ma

llms reinforcement-learning multimodal benchmarks

arXiv AI

6d ago

Stepwise Intrinsic Rewards for Reasoning in Large Language Models

The paper introduces Stepwise Marginal Information Gain (MIG), an intrinsic process reward that evaluates how each reasoning step of a large language model (LLM) or vision-language model (VLM) improves the likelihood of the reference answer. MIG rewards only new likelihood maxima, preventing duplicate credit, and is combined with outcome, format, and self‑distillation objectives to guide training. Experiments on eight benchmarks show that this method outperforms outcome‑only reinforcement learning and improves accuracy by up to 4.8 points over binary‑reward training, including a 12.6‑point gain on MathVerse and a 12.9‑point advantage on vision‑language transfer at 7B parameters.

By Xiangwei Wang, Wei Wang, Ken Chen, Nanduni Nimalsiri, Sachith Seneviratne, Saman Halgamuge

llms rag reinforcement-learning efficiency multimodal benchmarks

arXiv AI

Jun 11

ProcessThinker: Enhancing Multi-modal Large Language Models Reasoning via Rollout-based Process Reward

arXiv:2606. 11209v1 Announce Type: cross Abstract: Visual question answering increasingly requires multi-step reasoning.

By Jingpei Wu, Xiao Han, Weixiang Shen, Boer Zhang, Zifeng Ding, Volker Tresp

llms nlp reinforcement-learning fine-tuning multimodal benchmarks

arXiv Machine Learning

Jul 14

The Cost of Reasoning: Chain-of-Thought Induces Overconfidence in Vision-Language Models

arXiv:2603. 16728v2 Announce Type: replace Abstract: Vision-language models (VLMs) are increasingly deployed in high-stakes settings where reliable uncertainty quantification (UQ) is as important as predictive accuracy.

By Robert Welch, Emir Konuk, Kevin Smith

llms multimodal

One email a morning, machine-written

Related stories

Aligning Thoughts with Answers: Probability Rewards to Tame Thinking Drift

On Language Drift during RLVR Post-Training

Be Faithful When Response: Returning Fluent and Grounded Answers for Vision-Language Models Reinforcement Learning

Stepwise Intrinsic Rewards for Reasoning in Large Language Models

ProcessThinker: Enhancing Multi-modal Large Language Models Reasoning via Rollout-based Process Reward

The Cost of Reasoning: Chain-of-Thought Induces Overconfidence in Vision-Language Models