Skip to content
arXiv · Computation and Language· Natapong Nitarach, Phume Ngampornsukswadi, Pittawat Taveekitworachai, Surapon Nonesung, Warit Sirichotedumrong, Kunat Pipatanakul·· 3 hr agoAI score32

AMLworld-Compact:用于低成本 LLM 评测与错误诊断的重要性加权下采样

AMLworld-Compact: Importance-Weighted Downsampling for Cost-Effective LLM Evaluation and Error Diagnosis

AI brief

AMLworld-Compact 推出 HT-Coreset 协议,在 HI-Small 和 LI-Small 上将评测目标分别减少 271x 和 611x,保留全部非法交易并按监督分数分层抽样良性交易。

Source: arXiv · Computation and Language · arxiv.org