LLM Digest
Subscribe

Story

arxiv_cs_lg · Nov 17, 2026 · paper

Source brief

Learning Stratified Policies for Inference-Time Objectives

arxiv.orgNov 17, 2026
original source linked

In brief

Inference-time scaling improves the performance of large language models by sampling multiple responses and evaluating their collective utility. However, many existing approaches obtain these samples by repeatedly sampling from a single policy. In this work, we study if the sampl

Feed lens
eval

Continue reading

Read the original at arxiv.org →Open in live feedRead that day’s brief

Earlier in this thread 4 items