MOSS-ChatV: Reinforcement Learning with Process Reasoning Reward for Video Temporal Reasoning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tao, Sicheng, Li, Jungang, Yan, Yibo, Zhang, Junyan, Gao, Yubo, Li, Hanqian, Xun, ShuHang, Fan, Yuxuan, Chen, Hong, He, Jianxiang, Hu, Xuming
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915515858944000
author Tao, Sicheng
Li, Jungang
Yan, Yibo
Zhang, Junyan
Gao, Yubo
Li, Hanqian
Xun, ShuHang
Fan, Yuxuan
Chen, Hong
He, Jianxiang
Hu, Xuming
author_facet Tao, Sicheng
Li, Jungang
Yan, Yibo
Zhang, Junyan
Gao, Yubo
Li, Hanqian
Xun, ShuHang
Fan, Yuxuan
Chen, Hong
He, Jianxiang
Hu, Xuming
contents Video reasoning has emerged as a critical capability for multimodal large language models (MLLMs), requiring models to move beyond static perception toward coherent understanding of temporal dynamics in complex scenes. Yet existing MLLMs often exhibit process inconsistency, where intermediate reasoning drifts from video dynamics even when the final answer is correct, undermining interpretability and robustness. To address this issue, we introduce MOSS-ChatV, a reinforcement learning framework with a Dynamic Time Warping (DTW)-based process reward. This rule-based reward aligns reasoning traces with temporally grounded references, enabling efficient process supervision without auxiliary reward models. We further identify dynamic state prediction as a key measure of video reasoning and construct MOSS-Video, a benchmark with annotated reasoning traces, where the training split is used to fine-tune MOSS-ChatV and the held-out split is reserved for evaluation. MOSS-ChatV achieves 87.2\% on MOSS-Video (test) and improves performance on general video benchmarks such as MVBench and MMVU. The framework consistently yields gains across different architectures, including Qwen2.5-VL and Phi-2, confirming its broad applicability. Evaluations with GPT-4o-as-judge further show that MOSS-ChatV produces more consistent and stable reasoning traces.
format Preprint
id arxiv_https___arxiv_org_abs_2509_21113
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MOSS-ChatV: Reinforcement Learning with Process Reasoning Reward for Video Temporal Reasoning
Tao, Sicheng
Li, Jungang
Yan, Yibo
Zhang, Junyan
Gao, Yubo
Li, Hanqian
Xun, ShuHang
Fan, Yuxuan
Chen, Hong
He, Jianxiang
Hu, Xuming
Computer Vision and Pattern Recognition
Video reasoning has emerged as a critical capability for multimodal large language models (MLLMs), requiring models to move beyond static perception toward coherent understanding of temporal dynamics in complex scenes. Yet existing MLLMs often exhibit process inconsistency, where intermediate reasoning drifts from video dynamics even when the final answer is correct, undermining interpretability and robustness. To address this issue, we introduce MOSS-ChatV, a reinforcement learning framework with a Dynamic Time Warping (DTW)-based process reward. This rule-based reward aligns reasoning traces with temporally grounded references, enabling efficient process supervision without auxiliary reward models. We further identify dynamic state prediction as a key measure of video reasoning and construct MOSS-Video, a benchmark with annotated reasoning traces, where the training split is used to fine-tune MOSS-ChatV and the held-out split is reserved for evaluation. MOSS-ChatV achieves 87.2\% on MOSS-Video (test) and improves performance on general video benchmarks such as MVBench and MMVU. The framework consistently yields gains across different architectures, including Qwen2.5-VL and Phi-2, confirming its broad applicability. Evaluations with GPT-4o-as-judge further show that MOSS-ChatV produces more consistent and stable reasoning traces.
title MOSS-ChatV: Reinforcement Learning with Process Reasoning Reward for Video Temporal Reasoning
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.21113