2 papers
cs.LG2026
QuasiMoTTo: Quasi-Monte Carlo Test-Time Scaling
Michael Y. Li, Anthony Zhan, Kanishk Gandhi +2
Scaling inference compute, by generating many parallel attempts per problem, is a costly but reliable lever for improving language model capabilities. By default these attempts are…
cs.LG2025
Simple Policy Gradients for Reasoning with Diffusion Language Models
Anthony Zhan
Diffusion large language models (dLLMs) represent a promising alternative to autoregressive LLMs; however, the lack of effective post-training techniques, including reinforcement l…