Qwen3.5用强化学习学会在污染检索中挑有用证据，成功率从22%提到26%，还不掉通用分。
SelectBench是一个控制性基准和训练集，用于大语言模型选择性采纳检索结果中的证据。基于Qwen3.5-4B，使用DAPO强化学习方法，通过确定性规则奖励或语义法官进行后训练。在325样本的SelectBench-v2测试集上，严格成功率从原始检查点的22.46%提升至DAPO-Rule的25.54%和DAPO-DeepSeek的26.46%。两种策略均减少了禁止内容采纳，生成了更短更聚焦的回答，但提示注入跟随能力未改善。在MMLU和干净HotpotQA上未见明显退化，表明通用能力保持。
Reinforcement Learning for Large Language Model Selective Evidence Adoption from Contaminated Retrieval Results
Retrieval-augmented large language models frequently face contexts that interleave useful evidence with misleading statements or instruction-like content. Blanket refusal discards valid evidence, whereas uncritical adoption yields incorrect or unsafe answers. The ability to selectively adopt relevant information while rejecting deceptive or harmful content is therefore critical for reliable deployment in real-world retrieval settings. We introduce SelectBench, a controlled benchmark and training set for selective evidence adoption, and post-train Qwen3.5-4B directly with DAPO using either deterministic rule rewards or a frozen semantic judge. On the corrected 325-example SelectBench-v2 test set, strict success rises from 22.46% for the original checkpoint to 25.54% with DAPO-Rule and 26.46% with DAPO-DeepSeek. Both trained policies reduce forbidden-content adoption and produce shorter, more focused responses, yet prompt-injection following does not improve. The paired gains are modest and fail to survive Holm correction, suggesting that stronger reward shaping or additional training iterations may be needed for more robust gains. DAPO-DeepSeek exhibits no material degradation on MMLU or clean HotpotQA, indicating that the post-training procedure preserves general capabilities. These results demonstrate a directional improvement in selective evidence use, while identifying injection resistance and statistical robustness as important remaining challenges for future work.