Contribution-aware Token Compression for Efficient Video Understanding via Reinforcement Learning

2026-02-02

AI Summary

This work addresses the high computational cost of video large language models during inference, which stems from redundant video tokens. Existing compression approaches rely on attention scores but fail to explicitly model each token’s actual contribution to prediction outcomes. To overcome this limitation, the authors propose CaCoVID, an algorithm that, for the first time, directly optimizes token selection based on their contribution to the final prediction. Leveraging reinforcement learning, CaCoVID actively searches for the optimal token subset, while integrating combinatorial policy optimization and an online sampling mechanism to drastically reduce the search space and accelerate convergence. Experiments demonstrate that CaCoVID significantly lowers computational overhead across multiple video understanding benchmarks while maintaining or even improving model performance.

Technology Category

• Computer Vision: Large Vision Models
• Search and Optimization: Learning to Search
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Video large language models have demonstrated remarkable capabilities in video understanding tasks. However, the redundancy of video tokens introduces significant computational overhead during inference, limiting their practical deployment. Many compression algorithms are proposed to prioritize retaining features with the highest attention scores to minimize perturbations in attention computations. However, the correlation between attention scores and their actual contribution to correct answers remains ambiguous. To address the above limitation, we propose a novel \textbf{C}ontribution-\textbf{a}ware token \textbf{Co}mpression algorithm for \textbf{VID}eo understanding (\textbf{CaCoVID}) that explicitly optimizes the token selection policy based on the contribution of tokens to correct predictions. First, we introduce a reinforcement learning-based framework that optimizes a policy network to select video token combinations with the greatest contribution to correct predictions. This paradigm shifts the focus from passive token preservation to active discovery of optimal compressed token combinations. Secondly, we propose a combinatorial policy optimization algorithm with online combination space sampling, which dramatically reduces the exploration space for video token combinations and accelerates the convergence speed of policy optimization. Extensive experiments on diverse video understanding benchmarks demonstrate the effectiveness of CaCoVID. Codes will be released.

Problem

• video understanding
• token compression
• computational overhead
• attention scores
• contribution-aware

Innovation

• Contribution-aware
• Token Compression
• Reinforcement Learning
• Video Understanding
• Combinatorial Policy Optimization