Showing cs.LGShow all
2 papers · 1 filter
cs.LG2026
Same Answer, New Box: How Much of RLVR's Out-of-Domain Gain Is Answer Format?
Zongye Lyu
Reinforcement learning with verifiable rewards (RLVR) is credited with gains beyond the domain its rewards check, such as higher MMLU accuracy after training on mathematics. Such a…
cs.LG2026
Exact and Certified Data Shapley for Weighted Nearest Neighbor Regression
Zongye Lyu, Lihua Peng
Data Shapley can be computed exactly and efficiently for unweighted K-nearest-neighbor (KNN) models, the basis of the popular KNN-Shapley. For weighted KNN, however, efficient exac…