Benchmarking Multimodal CoT Reward Model Stepwise by Visual Program

• URL: http://arxiv.org/abs/2504.06606v1
• Date: Wed, 09 Apr 2025 06:09:40 GMT
• Title: Benchmarking Multimodal CoT Reward Model Stepwise by Visual Program
• Authors: Minghe Gao, Xuqi Liu, Zhongqi Yue, Yang Wu, Shuang Chen, Juncheng Li, Siliang Tang, Fei Wu, Tat-Seng Chua, Yueting Zhuang,
• Abstract summary: We propose SVIP, a novel approach to train a step-level multi-dimensional Chain-of-Thought(CoT) reward model automatically.<n>It generates code for solving visual tasks and transforms the analysis of code blocks into the evaluation of CoT step as training samples.<n>We show that SVIP-Reward improves MLLM performance across training and inference-time scaling.
• Score: 96.79600297158271
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent advancements in reward signal usage for Large Language Models (LLMs) are remarkable. However, significant challenges exist when transitioning reward signal to the multimodal domain, including labor-intensive annotations, over-reliance on one-step rewards, and inadequate evaluation. To address these issues, we propose SVIP, a novel approach to train a step-level multi-dimensional Chain-of-Thought~(CoT) reward model automatically. It generates code for solving visual tasks and transforms the analysis of code blocks into the evaluation of CoT step as training samples. Then, we train SVIP-Reward model using a multi-head attention mechanism called TriAtt-CoT. The advantages of SVIP-Reward are evident throughout the entire process of MLLM. We also introduce a benchmark for CoT reward model training and testing. Experimental results demonstrate that SVIP-Reward improves MLLM performance across training and inference-time scaling, yielding better results on benchmarks while reducing hallucinations and enhancing reasoning ability.

Related papers

• Boosting Virtual Agent Learning and Reasoning: A Step-wise, Multi-dimensional, and Generalist Reward Model with Benchmark [72.46357004059661]
  We propose Similar, a step-wise Multi-dimensional Generalist Reward Model. It offers fine-grained signals for agent training and can choose better action for inference-time scaling. We introduce the first benchmark in the virtual agent domain for step-wise, multi-dimensional reward model training and evaluation.
  arXiv  Detail & Related papers  (2025-03-24T13:30:47Z)
• Agentic Reward Modeling: Integrating Human Preferences with Verifiable Correctness Signals for Reliable Reward Systems [54.4392552373835]
  Reward models (RMs) are crucial for the training and inference-time scaling up of large language models (LLMs) We propose agentic reward modeling, a reward system that combines reward models with verifiable correctness signals to provide reliable rewards. We conduct comprehensive experiments on existing reward model benchmarks and inference time best-of-n searches on real-world downstream tasks.
  arXiv  Detail & Related papers  (2025-02-26T17:19:12Z)
• Utility-inspired Reward Transformations Improve Reinforcement Learning Training of Language Models [6.472081755630166]
  We show how linear aggregation of rewards exhibits some vulnerabilities.<n>We propose a transformation of reward functions inspired by economic theory of utility functions.<n>We show that models trained with Inada-transformations score as more helpful while being less harmful.
  arXiv  Detail & Related papers  (2025-01-08T19:03:17Z)
• Beyond Simple Sum of Delayed Rewards: Non-Markovian Reward Modeling for Reinforcement Learning [44.770495418026734]
  Reinforcement Learning (RL) empowers agents to acquire various skills by learning from reward signals. Traditional methods assume the existence of underlying Markovian rewards and that the observed delayed reward is simply the sum of instance-level rewards. We propose Composite Delayed Reward Transformer (CoDeTr), which incorporates a specialized in-sequence attention mechanism.
  arXiv  Detail & Related papers  (2024-10-26T13:12:27Z)
• A Large Language Model-Driven Reward Design Framework via Dynamic Feedback for Reinforcement Learning [25.82540393199001]
  CARD is a Reward Design framework that iteratively generates and improves reward function code. CARD includes a Coder that generates and verifies the code, while a Evaluator provides dynamic feedback to guide the Coder in improving the code.
  arXiv  Detail & Related papers  (2024-10-18T17:51:51Z)
• Step-by-Step Reasoning for Math Problems via Twisted Sequential Monte Carlo [55.452453947359736]
  We introduce a novel verification method based on Twisted Sequential Monte Carlo (TSMC)<n>We apply TSMC to Large Language Models by estimating the expected future rewards at partial solutions.<n>This approach results in a more straightforward training target that eliminates the need for step-wise human annotations.
  arXiv  Detail & Related papers  (2024-10-02T18:17:54Z)
• Routing to the Expert: Efficient Reward-guided Ensemble of Large Language Models [69.51130760097818]
  We propose Zooter, a reward-guided routing method distilling rewards on training queries to train a routing function. We evaluate Zooter on a comprehensive benchmark collection with 26 subsets on different domains and tasks.
  arXiv  Detail & Related papers  (2023-11-15T04:40:43Z)
• Let's Reinforce Step by Step [10.65244642965387]
  We use Reinforcement Learning from Human Feedback to shape model reasoning processes. Our results show that the fine-grained reward provided by PRM-based methods enhances accuracy on simple mathematical reasoning. We also show the critical role reward aggregation functions play in model performance.
  arXiv  Detail & Related papers  (2023-11-10T01:35:51Z)
• Let's reward step by step: Step-Level reward model as the Navigators for Reasoning [64.27898739929734]
  Process-Supervised Reward Model (PRM) furnishes LLMs with step-by-step feedback during the training phase. We propose a greedy search algorithm that employs the step-level feedback from PRM to optimize the reasoning pathways explored by LLMs. To explore the versatility of our approach, we develop a novel method to automatically generate step-level reward dataset for coding tasks and observed similar improved performance in the code generation tasks.
  arXiv  Detail & Related papers  (2023-10-16T05:21:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.