Embodied Reasoning on ALFWorld
0.96AccuracyStudent (IC+Cascade)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Student (IC+Cascade)Model=GPT-4.1-mini, In-Context Distillation=true, Self-Consistency Cascade=true2025.12 | 0.96 | 0.41 | |
| Student (IC+Cascade)Model=Llama-3.3-70B, In-Context Distillation=true, Self-Consistency Cascade=true2025.12 | 0.93 | 0.21 | |
| M2CLModel=Qwen-72B, Number of LLMs=642026.02 | 0.929 | — | |
| M2CLModel=Llama-70B, Number of LLMs=642026.02 | 0.908 | — | |
| TeacherModel=Sonnet 4.52025.12 | 0.89 | 1 | |
| Student (IC)Model=GPT-4.1-mini, In-Context Distillation=true2025.12 | 0.87 | 0.43 | |
| Student (IC)Model=Llama-3.3-70B, In-Context Distillation=true2025.12 | 0.87 | 0.14 | |
| M2CLBackbone=Qwen-72B, Number of LLMs=82026.02 | 0.79 | — | |
| M2CLBackbone Model=Qwen-72B, Number of LLMs=42026.02 | 0.79 | — | |
| M2CLBackbone=Llama-70B, Number of LLMs=82026.02 | 0.789 | — | |
| MacNetModel=Qwen-72B, Number of LLMs=642026.02 | 0.744 | — | |
| M2CLBase Model=Qwen-72B, Number of LLMs=42026.02 | 0.739 | — | |
| M2CLBase Model=Llama-70B, Number of LLMs=42026.02 | 0.695 | — | |
| MacNetModel=Llama-70B, Number of LLMs=642026.02 | 0.679 | — | |
| DebateModel=Qwen-72B, Number of LLMs=642026.02 | 0.625 | — | |
| MacNetBackbone=Qwen-72B, Number of LLMs=82026.02 | 0.618 | — | |
| MacNetBackbone Model=Qwen-72B, Number of LLMs=42026.02 | 0.618 | — | |
| DebateBackbone=Qwen-72B, Number of LLMs=82026.02 | 0.604 | — | |
| DebateBackbone Model=Qwen-72B, Number of LLMs=42026.02 | 0.604 | — | |
| BoNModel=Qwen-72B, Number of LLMs=642026.02 | 0.601 | — | |
| GPTSwarmModel=Qwen-72B, Number of LLMs=642026.02 | 0.596 | — | |
| DebateBase Model=Qwen-72B, Number of LLMs=42026.02 | 0.583 | — | |
| BoNBackbone=Qwen-72B, Number of LLMs=82026.02 | 0.575 | — | |
| Best-of-N (BoN)Backbone Model=Qwen-72B, Number of LLMs=42026.02 | 0.575 | — | |
| MacNetBackbone=Llama-70B, Number of LLMs=82026.02 | 0.573 | — | |
| MacNetBase Model=Qwen-72B, Number of LLMs=42026.02 | 0.572 | — | |
| DyLANModel=Qwen-72B, Number of LLMs=642026.02 | 0.569 | — | |
| GPTSwarmBackbone=Qwen-72B, Number of LLMs=82026.02 | 0.568 | — | |
| GPTSwarmBackbone Model=Qwen-72B, Number of LLMs=42026.02 | 0.568 | — | |
| BoNBase Model=Qwen-72B, Number of LLMs=42026.02 | 0.566 | — | |
| M2CLModel=Qwen-14B, Number of LLMs=642026.02 | 0.564 | — | |
| GPTSwarmBase Model=Qwen-72B, Number of LLMs=42026.02 | 0.552 | — | |
| DyLANBackbone=Qwen-72B, Number of LLMs=82026.02 | 0.551 | — | |
| DyLANBackbone Model=Qwen-72B, Number of LLMs=42026.02 | 0.551 | — | |
| M2CLModel=Llama-14B, Number of LLMs=642026.02 | 0.548 | — | |
| BoNModel=Llama-70B, Number of LLMs=642026.02 | 0.548 | — | |
| DyLANBase Model=Qwen-72B, Number of LLMs=42026.02 | 0.542 | — | |
| DebateModel=Llama-70B, Number of LLMs=642026.02 | 0.536 | — | |
| MacNetBase Model=Llama-70B, Number of LLMs=42026.02 | 0.527 | — | |
| BoNBackbone=Llama-70B, Number of LLMs=82026.02 | 0.524 | — | |
| GPTSwarmModel=Llama-70B, Number of LLMs=642026.02 | 0.524 | — | |
| DebateBackbone=Llama-70B, Number of LLMs=82026.02 | 0.517 | — | |
| GPTSwarmBackbone=Llama-70B, Number of LLMs=82026.02 | 0.514 | — | |
| GPTSwarmBase Model=Llama-70B, Number of LLMs=42026.02 | 0.51 | — | |
| BoNBase Model=Llama-70B, Number of LLMs=42026.02 | 0.509 | — | |
| DyLANModel=Llama-70B, Number of LLMs=642026.02 | 0.504 | — | |
| DebateBase Model=Llama-70B, Number of LLMs=42026.02 | 0.504 | — | |
| DyLANBackbone=Llama-70B, Number of LLMs=82026.02 | 0.502 | — | |
| Student (ZS)Model=Llama-3.3-70B, Zero-shot=true2025.12 | 0.5 | 0.11 | |
| DyLANBase Model=Llama-70B, Number of LLMs=42026.02 | 0.493 | — | |
| SingleBase Model=Qwen-72B, Number of LLMs=12026.02 | 0.493 | — | |
| M2CLBackbone=Qwen-14B, Number of LLMs=82026.02 | 0.482 | — | |
| SingleBackbone=Qwen-72B, Number of LLMs=82026.02 | 0.482 | — | |
| M2CLBackbone Model=Qwen-14B, Number of LLMs=42026.02 | 0.482 | — | |
| Single executionBackbone Model=Qwen-72B, Number of LLMs=42026.02 | 0.482 | — | |
| MacNetModel=Qwen-14B, Number of LLMs=642026.02 | 0.475 | — | |
| M2CLModel=Qwen-7B, Number of LLMs=642026.02 | 0.468 | — | |
| SingleModel=Qwen-72B, Number of LLMs=642026.02 | 0.467 | — | |
| M2CLBackbone=Llama-14B, Number of LLMs=82026.02 | 0.461 | — | |
| SingleBase Model=Llama-70B, Number of LLMs=12026.02 | 0.461 | — | |
| M2CLBase Model=Qwen-14B, Number of LLMs=42026.02 | 0.45 | — | |
| M2CLModel=Llama-7B, Number of LLMs=642026.02 | 0.442 | — | |
| SingleBackbone=Llama-70B, Number of LLMs=82026.02 | 0.441 | — | |
| SingleModel=Llama-70B, Number of LLMs=642026.02 | 0.43 | — | |
| M2CLBase Model=Llama-14B, Number of LLMs=42026.02 | 0.429 | — | |
| MacNetModel=Qwen-7B, Number of LLMs=642026.02 | 0.416 | — | |
| M2CLBackbone=Qwen-7B, Number of LLMs=82026.02 | 0.399 | — | |
| M2CLBackbone Model=Qwen-7B, Number of LLMs=42026.02 | 0.399 | — | |
| MacNetBackbone=Qwen-14B, Number of LLMs=82026.02 | 0.392 | — | |
| MacNetBackbone Model=Qwen-14B, Number of LLMs=42026.02 | 0.392 | — | |
| MacNetModel=Llama-14B, Number of LLMs=642026.02 | 0.391 | — | |
| MacNetModel=Llama-7B, Number of LLMs=642026.02 | 0.388 | — | |
| DebateModel=Qwen-14B, Number of LLMs=642026.02 | 0.388 | — | |
| BoNModel=Qwen-14B, Number of LLMs=642026.02 | 0.384 | — | |
| BoNBackbone=Qwen-14B, Number of LLMs=82026.02 | 0.375 | — | |
| Best-of-N (BoN)Backbone Model=Qwen-14B, Number of LLMs=42026.02 | 0.375 | — | |
| M2CLBackbone=Llama-7B, Number of LLMs=82026.02 | 0.372 | — | |
| DebateBackbone=Qwen-14B, Number of LLMs=82026.02 | 0.368 | — | |
| DebateBackbone Model=Qwen-14B, Number of LLMs=42026.02 | 0.368 | — | |
| M2CLBase Model=Qwen-7B, Number of LLMs=42026.02 | 0.367 | — | |
| DyLANModel=Qwen-14B, Number of LLMs=642026.02 | 0.362 | — | |
| BoNBase Model=Qwen-14B, Number of LLMs=42026.02 | 0.362 | — | |
| MacNetBase Model=Qwen-14B, Number of LLMs=42026.02 | 0.361 | — | |
| GPTSwarmModel=Qwen-14B, Number of LLMs=642026.02 | 0.359 | — | |
| DebateBase Model=Qwen-14B, Number of LLMs=42026.02 | 0.356 | — | |
| M2CLBase Model=Llama-7B, Number of LLMs=42026.02 | 0.351 | — | |
| GPTSwarmBackbone=Qwen-14B, Number of LLMs=82026.02 | 0.349 | — | |
| GPTSwarmBackbone Model=Qwen-14B, Number of LLMs=42026.02 | 0.349 | — | |
| DyLANBackbone=Qwen-14B, Number of LLMs=82026.02 | 0.348 | — | |
| DyLANBackbone Model=Qwen-14B, Number of LLMs=42026.02 | 0.348 | — | |
| GPTSwarmBase Model=Qwen-14B, Number of LLMs=42026.02 | 0.346 | — | |
| BoNModel=Qwen-7B, Number of LLMs=642026.02 | 0.345 | — | |
| MacNetBackbone=Qwen-7B, Number of LLMs=82026.02 | 0.336 | — | |
| MacNetBackbone Model=Qwen-7B, Number of LLMs=42026.02 | 0.336 | — | |
| DyLANBase Model=Qwen-14B, Number of LLMs=42026.02 | 0.336 | — | |
| MacNetBackbone=Llama-14B, Number of LLMs=82026.02 | 0.334 | — | |
| BoNModel=Llama-14B, Number of LLMs=642026.02 | 0.326 | — | |
| DebateModel=Llama-14B, Number of LLMs=642026.02 | 0.326 | — | |
| DebateModel=Qwen-7B, Number of LLMs=642026.02 | 0.323 | — | |
| BoNBackbone=Llama-14B, Number of LLMs=82026.02 | 0.319 | — |