LLM Digest

Story

arxiv_llm_reliability · Aug 31, 2026 · paper

Source brief

Lazy Grounding: Attacking Search Agents with Factual Evidence

arxiv.orgAug 31, 2026
original source linked

In brief

Search agents reduce hallucination by grounding answers in retrieved web evidence. Yet reliance on retrieval also creates an attack surface: poisoned corpora with false or malicious documents can cause agents to repro...

Feed lens

agenteval

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

Lazy Grounding: Attacking Search Agents with Factual Evidence

Earlier in this thread 4 items

Teaching Nemotron Greek: Mining a Corpus, Adapting Retrieval, and Grounding Generation for Modern Greek across Specialist Domains

Abduction Without a Body? Representational Grounding and the Abduction Loop for Scientific Hypothesis Generation

Can AI agents conduct open-ended AI research? Early evidence from two case studies

OmegaUse-OfficeVal: Benchmarking LLM Agents on Long-Horizon Office-Suite Tasks with Economic Grounding