ReadingResearchRadarInvestment framework
Sign in / Sign up
← Back to reading
NewsTopic · 懒惰接地攻击研究2 min read·2026-10-15

Study Finds Injecting Factual But Distracting Evidence Can Drop Search Agent Accuracy by Average 5.9 Points

New research reveals 'lazy grounding': injecting real but distracting documents supporting neighboring questions into search corpora causes search agents to adopt incorrect answers, dropping accuracy by an average of 5.9 points and up to 17.3 points across 12 model-benchmark pairs.

ImportanceMaterialEvidenceE2 unreplicatedWrite-upQuick

2026-10-15 12:00 GMT+8

Researchers have identified a failure mode called "lazy grounding," demonstrating that search agents can be misled by completely factual but distracting information, not just misinformation. The study, authored by Yulin Zhang et al., was submitted to arXiv on August 31, 2026, revised to v3 on October 13, 2026, and accepted to the EMNLP 2026 Main Conference.

While retrieval-based approaches mitigate hallucination by grounding answers in web results, they introduce an attack surface. Previous concerns focused on poisoned corpora containing false or malicious documents. This work shows that even truthful documents can degrade performance if they support a neighboring rewritten question rather than the original query, leading agents to misapply valid evidence.

The team exposed lazy grounding by injecting nearby evidence from answer-changing rewrites of benchmark questions into search corpora. Each injected document contained factual evidence supporting a neighboring rewritten question but was retrieved for the original question. Across 12 model-benchmark pairs, the attack caused the accuracy of search agents' responses to drop by 5.9 points on average and by up to 17.3 points, inducing nearby-answer adoption in every setting. The effect was stronger when the distracting evidence appeared later in the retrieval list or was more answer-shaped.

These results indicate that robust search agents must defend against the misapplication of factual evidence, not just misinformation. Code is publicly available at github.com/frankyzha/lazy-grounding. These are first-party benchmark results; independent reproduction has not yet been reported.

Sign in / Sign up
ReadingResearchRadarInvestment framework

Sources

  • 原始出处arxiv.org↗
Topic

懒惰接地攻击研究

杜克大学团队在EMNLP 2026发表研究,揭示“懒惰接地”(Lazy Grounding)攻击:向搜索语料库注入支持邻近问题的真实但干扰性文档,导致搜索智能体准确率平均下降5.9个百分点。该攻击无需伪造信息,仅利用事实证据的误用即可误导模…

2 steps0 piecesUpdated 2026-10-16 01:38

How it got here

  1. 2026-08-31论文《Lazy Grounding: Attacking Search Agents with Factual Evidence》首次提交至arXiv。
  2. 2026-10-13论文更新至v3版本,确认被EMNLP 2026主会议接收,并公开代码仓库。

What is still unsettled

  • 主流搜索智能体厂商是否会针对“事实证据误用”更新其检索增强生成(RAG)的安全过滤策略?
  • 在实际生产环境中,如何区分“邻近问题的正确答案”与“当前问题的错误答案”而不显著增加推理延迟?
Read the full topic →
Continue researching →