value-consistency

Tag

Cards List
#value-consistency

VALCON: Benchmarking and Enhancing Value Consistency in Large Language Models via Dual-Role Variation

arXiv cs.CL ↗ · 2d ago Cached

该论文提出VALCON框架,通过生成者/评判者双重角色变体来基准测试并提升大语言模型的价值一致性,实验揭示28个LLM普遍存在跨角色不一致问题,并提出基于表征的干预方法加以缓解。

0 favorites 0 likes
← Back to home

Submit Feedback