按年被提到
20252025 · 2 次2026
大模型与语言模型›大模型评测与真实性·arXiv:2503.11751 · cs.CL

reWordBench: Benchmarking and Improving the Robustness of Reward Models with Transformed Inputs

Zhaofeng Wu, Michihiro Yasunaga, Andrew Cohen, Yoon Kim, Asli Celikyilmaz, Marjan Ghazvininejad · 2025
开始阅读 ↗PDFalphaXiv 讨论
被提及 · 2 人首次 2025-03-18 · 最近 2025-04-01
2不同的人
2活跃月份
2讨论串
—被引(OpenAlex)
arXiv 摘要(候选层不写导读)

Reward models have become a staple in modern NLP, serving as not only a scalable text evaluator, but also an indispensable component in many alignment recipes and inference-time algorithms. However, while recent reward models increase performance on standard benchmarks, this may partly be due to overfitting effects, which would confound an understanding of their true capability. In this work, we scrutinize the robustness of reward models and the extent of such overfitting. We build **reWordBench**, which systematically transforms reward model inputs in meaning- or ranking-preserving ways. We show that state-of-the-art reward models suffer from substantial performance degradation even with minor input transformations, sometimes dropping to significantly below-random accuracy, suggesting brittleness. To improve reward model robustness, we propose to explicitly train them to assign similar scores to paraphrases, and find that this approach also improves robustness to other distinct kinds of transformations. For example, our robust reward model reduces such degradation by roughly half for the Chat Hard subset in RewardBench. Furthermore, when used in alignment, our robust reward models demonstrate better utility and lead to higher-quality outputs, winning in up to 59% of instances against a standardly trained RM.

候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。

HN 书评

没有评论直接链到它;下面是它被提交的讨论串里回复最多的评论(不计入统计)

讨论串里没有 ≥100 字的评论,只有提交记录。

被提交为 story · 2 次