List your company

Embodied Task Planning on VirtualHome (unseen domains)

80.16Success RateTMoW
4.406424.073243.7463.4068Jan 30, 2026

Evaluation results

MethodLinks
TMoWBackbone=Llama-3.2-1B2026.0180.1613.2
SayCanPaySay Backbone=Llama-3.2-3B, Pay Backbone=Llama-3.2-1B2026.0149.5318.55
LLM+FTBackbone=Llama-3.2-1B, Training=Fine-tuned2026.0144.2421
FLAREBackbone=Llama-3.2-3B2026.0140.0722.57
LLM-PlannerBackbone=Llama-3.2-3B, Adaptation=Few-shot2026.0136.0522.93
ZSPBackbone=Llama-3.2-3B, Adaptation=Zero-shot2026.017.3228.22