List your company

Question Answering on 2WikiMultihopQA (EM/ASR)

60EMClean (No Attack)
6.9620.7334.548.27Jul 1, 2026

Evaluation results

MethodLinks
Clean (No Attack)Framework=ReAct, LLM Backbone=Qwen2.5-32B-Inst2026.0760—
IgnoreFramework=WebThinker, LLM Backbone=QwQ-32B2026.07591
IgnoreFramework=WebThinker, LLM Backbone=DeepSeek-32B2026.07590
Clean (No Attack)Framework=WebThinker, LLM Backbone=DeepSeek-32B2026.0758—
CombinedFramework=WebThinker, LLM Backbone=DeepSeek-32B2026.07580
TopicFramework=WebThinker, LLM Backbone=DeepSeek-32B2026.07580
Clean (No Attack)Framework=WebThinker, LLM Backbone=QwQ-32B2026.0757—
NaiveFramework=WebThinker, LLM Backbone=QwQ-32B2026.07563
NaiveFramework=WebThinker, LLM Backbone=DeepSeek-32B2026.07550
Clean (No Attack)Framework=ReAct, LLM Backbone=Llama-3.3-70B-Inst2026.0754—
TopicFramework=WebThinker, LLM Backbone=QwQ-32B2026.07540
CombinedFramework=WebThinker, LLM Backbone=QwQ-32B2026.07533
PARADOXFramework=WebThinker, LLM Backbone=QwQ-32B2026.07513
Fake CompletionFramework=WebThinker, LLM Backbone=DeepSeek-32B2026.07511
Fake CompletionFramework=ReAct, LLM Backbone=Llama-3.3-70B-Inst2026.07491
Fake CompletionFramework=WebThinker, LLM Backbone=QwQ-32B2026.07488
NaiveFramework=ReAct, LLM Backbone=Llama-3.3-70B-Inst2026.074710
PoisonedRAGFramework=ReAct, LLM Backbone=Llama-3.3-70B-Inst2026.07410
PARADOXFramework=ReAct, LLM Backbone=Llama-3.3-70B-Inst2026.074115
PARADOXFramework=WebThinker, LLM Backbone=DeepSeek-32B2026.073810
NaiveFramework=ReAct, LLM Backbone=Qwen2.5-32B-Inst2026.073733
PoisonedRAGFramework=ReAct, LLM Backbone=Qwen2.5-32B-Inst2026.07372
PARADOXFramework=ReAct, LLM Backbone=Qwen2.5-32B-Inst2026.073621
PoisonedRAGFramework=WebThinker, LLM Backbone=DeepSeek-32B2026.07352
IgnoreFramework=ReAct, LLM Backbone=Qwen2.5-32B-Inst2026.073417
CombinedFramework=ReAct, LLM Backbone=Llama-3.3-70B-Inst2026.07341
PoisonedRAGFramework=WebThinker, LLM Backbone=QwQ-32B2026.07342
KidnapRAGFramework=WebThinker, LLM Backbone=QwQ-32B2026.072946
CombinedFramework=ReAct, LLM Backbone=Qwen2.5-32B-Inst2026.072637
Fake CompletionFramework=ReAct, LLM Backbone=Qwen2.5-32B-Inst2026.072453
IgnoreFramework=ReAct, LLM Backbone=Llama-3.3-70B-Inst2026.072418
KidnapRAGFramework=WebThinker, LLM Backbone=DeepSeek-32B2026.072039
TopicFramework=ReAct, LLM Backbone=Llama-3.3-70B-Inst2026.07181
TopicFramework=ReAct, LLM Backbone=Qwen2.5-32B-Inst2026.07157
KidnapRAGFramework=ReAct, LLM Backbone=Qwen2.5-32B-Inst2026.07954
KidnapRAGFramework=ReAct, LLM Backbone=Llama-3.3-70B-Inst2026.07962