Score: 1

Group-Aware Reinforcement Learning for Output Diversity in Large Language Models

Published: November 16, 2025 | arXiv ID: 2511.12596v1

By: Oron Anschel , Alon Shoshan , Adam Botach and more

BigTech Affiliations: Amazon

Potential Business Impact:

Makes AI give more different and interesting answers.

Business Areas:
Natural Language Processing Artificial Intelligence, Data and Analytics, Software

Large Language Models (LLMs) often suffer from mode collapse, repeatedly generating the same few completions even when many valid answers exist, limiting their diversity across a wide range of tasks. We introduce Group-Aware Policy Optimization (GAPO), a simple extension of the recent and popular Group Relative Policy Optimization (GRPO) that computes rewards over the group as a whole. GAPO enables learning from the group-level properties such as diversity and coverage. We demonstrate GAPO using a frequency-aware reward function that encourages uniform sampling over valid LLM completions, and show that GAPO-trained models produce valid and more diverse model responses. Beyond this setup, GAPO generalizes to open-ended prompts and improves response diversity without compromising accuracy on standard LLM benchmarks (GSM8K, MATH, HumanEval, MMLU-Pro). Our code will be made publicly available.

Country of Origin
🇺🇸 United States

Page Count
22 pages

Category
Computer Science:
Computation and Language