‹ back to 2026-07-22

Reinforcement Learning for Large Language Model Selective Evidence Adoption from Contaminated Retrieval Results

arXiv cs.CL LLM Agents Yanyu Chen, Yue Li, Yongyi Cui, Dongsheng Shi, Lichang Dai 2026-07-22

TL;DR - SelectBench trains LLMs to use valid retrieved evidence while rejecting misleading or harmful content. Reinforcement learning produced modest gains without degrading general capabilities, but did not improve prompt-injection resistance.

SelectBench-v2 evaluates selective evidence adoption across 325 corrected test examples.
DAPO post-training raised strict success from 22.46% to 25.54% with rule rewards and 26.46% with a frozen semantic judge.
Trained models adopted less forbidden content and generated shorter, more focused answers.
Gains did not survive Holm correction; MMLU and clean HotpotQA performance remained stable.

view merged work →