List your company

DeepSearch Question Answering on Xbench

34Average ScoreAPPO
6.9613.982128.02Jun 10, 2026

Evaluation results

MethodLinks
APPOBackbone=Qwen3-14B, Method Strategy=APPO, Method Category=RL-based Method2026.0634
DeepSeek-R1-671BBackbone=DeepSeek-R1-671B, Method Category=Reference2026.0632.7
ARPOBackbone=Qwen3-14B, Method Strategy=ARPO, Method Category=RL-based Method2026.0632
APPOBackbone=Qwen3-8B, Method Strategy=APPO, Method Category=RL-based Method2026.0628
GRPOBackbone=Qwen3-14B, Method Strategy=GRPO, Method Category=RL-based Method2026.0627
ARPOBackbone=Qwen3-8B, Method Strategy=ARPO, Method Category=RL-based Method2026.0625
WebThinkerBackbone=Qwen3-14B, Method Strategy=WebThinker, Method Category=Single-Enhanced Method2026.0623
Search-o1Backbone=Qwen3-14B, Method Strategy=Search-o1, Method Category=Single-Enhanced Method2026.0621
GRPOBackbone=Qwen3-8B, Method Strategy=GRPO, Method Category=RL-based Method2026.0620
ReActBackbone=Qwen3-14B, Method Strategy=ReAct, Method Category=Single-Enhanced Method2026.0620
GPT-4oBackbone=GPT-4o, Method Category=Reference2026.0618
ReActBackbone=Qwen3-8B, Method Strategy=ReAct, Method Category=Single-Enhanced Method2026.0616
Vanilla RAGBackbone=Qwen3-14B, Method Strategy=Vanilla RAG, Method Category=Single-Enhanced Method2026.0615
Qwen3-32B-thinkingBackbone=Qwen3-32B, Method Category=Reference2026.0614
Qwen3-14BBackbone=Qwen3-14B, Method Strategy=Base, Method Category=RL-based Method2026.0614
WebThinkerBackbone=Qwen3-8B, Method Strategy=WebThinker, Method Category=Single-Enhanced Method2026.0613
QwQ-32BBackbone=QwQ-32B, Method Category=Reference2026.0610.7
DeepSeek-R1-32BBackbone=DeepSeek-R1-32B, Method Category=Reference2026.0610
Search-o1Backbone=Qwen3-8B, Method Strategy=Search-o1, Method Category=Single-Enhanced Method2026.0610
Qwen3-8BBackbone=Qwen3-8B, Method Strategy=Base, Method Category=RL-based Method2026.069
Vanilla RAGBackbone=Qwen3-8B, Method Strategy=Vanilla RAG, Method Category=Single-Enhanced Method2026.068