Balancing Rewards in Text Summarization: Multi-Objective Reinforcement Learning via HyperVolume Optimization
This work addresses the challenge of generating balanced summaries across multiple quality dimensions for text summarization applications, representing an incremental improvement over existing methods like GRPO.
The paper tackles the multi-objective optimization problem in text summarization by introducing hypervolume optimization (HVO) to dynamically balance rewards for consistency, coherence, relevance, and fluency, resulting in a 7B model enhanced by HVO performing comparably to GPT-4 while maintaining shorter generation lengths.
Text summarization is a crucial task that requires the simultaneous optimization of multiple objectives, including consistency, coherence, relevance, and fluency, which presents considerable challenges. Although large language models (LLMs) have demonstrated remarkable performance, enhanced by reinforcement learning (RL), few studies have focused on optimizing the multi-objective problem of summarization through RL based on LLMs. In this paper, we introduce hypervolume optimization (HVO), a novel optimization strategy that dynamically adjusts the scores between groups during the reward process in RL by using the hypervolume method. This method guides the model's optimization to progressively approximate the pareto front, thereby generating balanced summaries across multiple objectives. Experimental results on several representative summarization datasets demonstrate that our method outperforms group relative policy optimization (GRPO) in overall scores and shows more balanced performance across different dimensions. Moreover, a 7B foundation model enhanced by HVO performs comparably to GPT-4 in the summarization task, while maintaining a shorter generation length. Our code is publicly available at https://github.com/ai4business-LiAuto/HVO.git