Optimising Factual Consistency in Summarisation via Preference Learning from Multiple Imperfect Metrics 文章

ArXiv CS.CL2026-05-27NEWSen作者: Yuxuan Ye, Raul Santos-Rodriguez, Edwin Simpson

查看原文 →

Optimising Factual Consistency in Summarisation via Preference Learning from Multiple Imperfect Metrics · 相关技术

相关技术

ODE language model 远程代码执行（RCE）reinforcement learning eval Verse Straight-Through Estimator PPR Narrative Abstraction Benchmark NAT MIT HIS For Effort Metric Attention ENA Coding Camouflaged object detection ANN API