AI Paper Atlas

Research map / Large language models / Agents, code and math

SafeSearch: Automated Red-Teaming of LLM-Based Search Agents

Jianshuo Dong, Sheng Guo, Hao Wang, Xun Chen, Zhuotao Liu, Tianwei Zhang, Ke Xu, Minlie Huang, Han Qiu

ICML 2026 · Large language models › Agents, code and math · approach: Software engineering agents

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
switch from search workflow to tool calling scaffoldGPT-4.1-mini · SafeSearchoverall ASR (%): 90.5 -> 77.8Improved
switch from search workflow to deep research scaffoldGPT-4.1-mini · SafeSearchoverall ASR (%): 90.5 -> 57.4Improved
force the tool-calling agent to use the full search budget (9 searches) (ablation)GPT-4.1-mini tool-calling · SafeSearchASR (%): 74.33 -> 36.67Improved
force deep-research to use the full search budget (9 searches) (ablation)GPT-4.1-mini deep research · SafeSearchASR (%): 57.00 -> 46.16Improved
increase forced search budget from 3 to 9 for tool calling (ablation)GPT-4.1-mini tool-calling · SafeSearchASR (%): 49.89 -> 36.67Improved
increase auto search budget from 3 to 9 for tool callingGPT-4.1-mini tool-calling · SafeSearchASR (%): 74.56 -> 74.33No clear effect
use reasoning models instead of non-reasoning onessearch agents · SafeSearchASRImproved
use newer-generation GPT-5 / Claude-4.5 backends instead of GPT-4.1-minisearch agents (tool calling) · SafeSearchoverall ASR (%): 77.8 -> 4.6 (Claude-Sonnet-4.5)Improved
provide fewer search results to the LLM (ablation)search workflow agents · SafeSearchASRHurt
add reminder prompting about unreliable websitessearch agents · SafeSearchASRNo clear effect
add an auxiliary filtering detector (GPT-4.1-mini) over search resultssearch agents · SafeSearchASRImproved
generate test cases against stronger baseline agent (GPT-5-mini instead of Qwen3-8B)SafeSearch test generation · SafeSearchASRImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy