List your company

Multi-hop tool use on ToolHop

53.97Answer CorrectnessClaude 4.0-Sonnet
10.310821.645432.9844.3146Aug 12, 2025Oct 2, 2025Nov 22, 2025Jan 12, 2026Mar 4, 2026Apr 24, 2026Jun 15, 2026

Evaluation results

MethodLinks
Claude 4.0-SonnetVersion=4.0-Sonnet2025.0853.97
Qwen3-32B ReasoningFamily=Qwen3-32B, Version=Reasoning2025.0850.12
PACTBackbone=Qwen3-4B2026.0649.55
MatchTIRBackbone=Qwen3-4B2026.0647.54
SFT→MatchTIRBackbone=Qwen3-4B2026.0647.34
MatchTIR (KM)Backbone=Qwen3-8B2026.0146.16
MatchTIR (OT)Backbone=Qwen3-8B2026.0145.8
Gemini 2.5-ProVersion=2.5-Pro2025.0845.32
Qwen3-14B Reasoning FTRL-Reinforce++Family=Qwen3-14B, Version=Reasoning, Method=FTRL-Reinforce++2025.0844.79
FTRL-MBackbone=Qwen3-8B2026.0143.32
FTRLBackbone=Qwen3-4B2026.0643.02
MatchTIR (KM)Backbone=Qwen3-4B2026.0142.55
ToolRL-MBackbone=Qwen3-8B2026.0142.55
VanillaBackbone=Qwen3-8B2026.0142.21
MatchTIR (OT)Backbone=Qwen3-4B2026.0141.95
Qwen3-14B Reasoning FTRL-GRPOFamily=Qwen3-14B, Version=Reasoning, Method=FTRL-GRPO2025.0841.64
FTRL-MBackbone=Qwen3-4B2026.0141.24
Qwen3-8B Reasoning FTRL-Reinforce++Family=Qwen3-8B, Version=Reasoning, Method=FTRL-Reinforce++2025.0841.24
Qwen3-8B ReasoningFamily=Qwen3-8B, Version=Reasoning2025.0840.7
GRPOBackbone=Qwen3-8B2026.0140.64
Qwen3-8B Reasoning FTRL-GRPOFamily=Qwen3-8B, Version=Reasoning, Method=FTRL-GRPO2025.0840.5
Qwen3-14B ReasoningFamily=Qwen3-14B, Version=Reasoning2025.0840.1
Gemini 2.5-FlashVersion=2.5-Flash2025.0839.37
Qwen3-8B Non-Reasoning FTRL-GRPOFamily=Qwen3-8B, Version=Non-Reasoning, Method=FTRL-GRPO2025.0838.93
FTRL-SBackbone=Qwen3-4B2026.0138.63
Qwen3-14B Non-Reasoning FTRL-Reinforce++Family=Qwen3-14B, Version=Non-Reasoning, Method=FTRL-Reinforce++2025.0838.32
Qwen3-14B Non-Reasoning FTRL-GRPOFamily=Qwen3-14B, Version=Non-Reasoning, Method=FTRL-GRPO2025.0838.26
Qwen3-8B Non-Reasoning FTRL-Reinforce++Family=Qwen3-8B, Version=Non-Reasoning, Method=FTRL-Reinforce++2025.0837.79
CHORDBackbone=Qwen3-4B2026.0637.69
GRPOBackbone=Qwen3-4B2026.0137.25
FTRL-SBackbone=Qwen3-8B2026.0136.72
Qwen2.5-14B FTRL-Reinforce++Family=Qwen2.5-14B, Method=FTRL-Reinforce++2025.0836.42
GPT 4oVersion=4o2025.0836.31
ToolRL-MBackbone=Qwen3-4B2026.0135.68
Qwen3-32B Non-ReasoningFamily=Qwen3-32B, Version=Non-Reasoning2025.0835.34
PACTBackbone=Qwen3-1.7B2026.0634.47
GRPOBackbone=Qwen3-4B2026.0634.47
ToolRLBackbone=Qwen3-4B2026.0634.27
ToolRL-SBackbone=Qwen3-8B2026.0132.93
MatchTIRBackbone=Qwen3-1.7B2026.0632.26
VanillaBackbone=Qwen3-4B2026.0631.86
VanillaBackbone=Qwen3-4B2026.0131.63
SFT→MatchTIRBackbone=Qwen3-1.7B2026.0631.06
Qwen2.5-72B InstructFamily=Qwen2.5-72B, Version=Instruct2025.0830.29
ToolRL-SBackbone=Qwen3-4B2026.0130.28
SFTBackbone=Qwen3-4B2026.0630.25
Qwen2.5-7B FTRL-GRPOFamily=Qwen2.5-7B, Method=FTRL-GRPO2025.0829.68
Qwen3-8B Non-ReasoningFamily=Qwen3-8B, Version=Non-Reasoning2025.0828.54
CHORDBackbone=Qwen3-1.7B2026.0627.14
Qwen2.5-14B FTRL-GRPOFamily=Qwen2.5-14B, Method=FTRL-GRPO2025.0826.63
FTRLBackbone=Qwen3-1.7B2026.0626.43
ToolRLBackbone=Qwen3-1.7B2026.0625.83
Qwen2.5-14B InstructFamily=Qwen2.5-14B, Version=Instruct2025.0825.8
GPT 3.5-TurboVersion=3.5-Turbo2025.0825.73
VanillaBackbone=Qwen3-1.7B2026.0625.63
GRPOBackbone=Qwen3-1.7B2026.0625.13
Qwen3-14B Non-ReasoningFamily=Qwen3-14B, Version=Non-Reasoning2025.0824.19
Qwen2.5-7B FTRL-Reinforce++Family=Qwen2.5-7B, Method=FTRL-Reinforce++2025.0823.45
SFTBackbone=Qwen3-1.7B2026.0623.32
Qwen2.5-32B InstructFamily=Qwen2.5-32B, Version=Instruct2025.0819.7
Qwen2.5-7B InstructFamily=Qwen2.5-7B, Version=Instruct2025.0811.99