List your company

Embodied Task Planning on Robotouille Asynchronous (test)

86Pass@1 AccuracyGiG+Exp
-3.4419.784366.22Jan 29, 2026

Evaluation results

MethodLinks
GiG+ExpBackbone=DeepSeek, Planning Framework=GiG+Exp, Experience Memory=Yes2026.0186
GiG+ExpBackbone=Qwen3, Planning Framework=GiG+Exp, Experience Memory=Yes2026.0182
GiGBackbone=Qwen3, Planning Framework=GiG2026.0172
GiGBackbone=Gemini, Planning Framework=GiG2026.0166
GiG+ExpBackbone=Gemini, Planning Framework=GiG+Exp, Experience Memory=Yes2026.0166
ReActBackbone=Gemini, Planning Framework=ReAct2026.0160
GiGBackbone=DeepSeek, Planning Framework=GiG2026.0159
ReCAPBackbone=Qwen3, Planning Framework=ReCAP2026.0135
ReActBackbone=Qwen3, Planning Framework=ReAct2026.0131
ReCAPBackbone=DeepSeek, Planning Framework=ReCAP2026.0127
ReCAPBackbone=Gemini, Planning Framework=ReCAP2026.0121
ReActBackbone=DeepSeek, Planning Framework=ReAct2026.0116
CoTBackbone=Gemini, Planning Framework=CoT2026.014
CoTBackbone=Qwen3, Planning Framework=CoT2026.010
CoTBackbone=DeepSeek, Planning Framework=CoT2026.010