List your company

Interactive Reasoning on ScienceWorld (Seen)

63.91Success RateITPR
-0.018816.578133.17549.7719Jan 13, 2026Jan 16, 2026Jan 20, 2026Jan 23, 2026Jan 27, 2026Jan 30, 2026Feb 3, 2026

Evaluation results

MethodLinks
ITPRBackbone=Llama3.1-8B, Type=Training2026.0163.91
ITPRBackbone=Qwen2.5-7B, Type=Training2026.0162.58
ITPRBackbone=Qwen3-8B, Type=Training2026.0161.85
AECBackbone=Llama3-8B2026.0261.36
WKMBackbone=Llama3.1-8B, Type=Training2026.0161.34
IWMBackbone=Qwen2.5-7B, Type=Training2026.0160.82
WKMBackbone=Qwen3-8B, Type=Training2026.0160.31
WKMBackbone=Llama3-8B2026.0260.12
IWMBackbone=Qwen3-8B, Type=Training2026.0159.27
KnowAgentBackbone=Llama3-8B2026.0258.67
ETOBackbone=Llama3-8B2026.0257.9
IWMBackbone=Llama3.1-8B, Type=Training2026.0157.56
SFTBackbone=Llama3.1-8B, Type=Training2026.0157.21
SFTBackbone=Qwen3-8B, Type=Training2026.0156.7
SFTBackbone=Qwen2.5-7B, Type=Training2026.0155.67
NATBackbone=Llama3-8B2026.0255.24
WKMBackbone=Qwen2.5-7B, Type=Training2026.0154.12
ReflexionBackbone=Llama3-8B2026.0227.23
ReActBackbone=Llama3-8B2026.0224.76
ITPIBackbone=Qwen3-8B, Type=Prompting2026.0120.61
ITPIBackbone=Llama3.1-8B, Type=Prompting2026.0119.58
ITPIBackbone=Qwen2.5-7B, Type=Prompting2026.0116.49
RAPBackbone=Qwen3-8B, Type=Prompting2026.0115.46
RAPBackbone=Llama3.1-8B, Type=Prompting2026.0111.34
RAPBackbone=Qwen2.5-7B, Type=Prompting2026.0110.3
ReActBackbone=Qwen3-8B, Type=Prompting2026.019.79
ReActBackbone=Llama3.1-8B, Type=Prompting2026.019.27
ReActBackbone=Qwen2.5-7B, Type=Prompting2026.018.24
CoTBackbone=Qwen2.5-7B, Type=Prompting2026.013.09
CoTBackbone=Llama3.1-8B, Type=Prompting2026.013.09
CoTBackbone=Qwen3-8B, Type=Prompting2026.012.44