List your company

Embodied Planning on ALFWorld Seen v1 (in-distribution)

57.14Average Success RateProCeedSFT
9.227221.666134.10546.5439Mar 9, 2025May 12, 2025Jul 16, 2025Sep 19, 2025Nov 23, 2025Jan 27, 2026Apr 2, 2026

Evaluation results

MethodLinks
ProCeedSFTBackbone=Qwen3-8B2026.0457.14———————
ProCeedRLBackbone=Qwen3-8B2026.0451.43———————
RFTBackbone=Qwen3-8B2026.0447.38———————
DAPOBackbone=Qwen3-8B2026.0445.23———————
Qwen3-8BBackbone=Qwen3-8B2026.0444.22———————
GFlowVLM w/ SubTBAssump.=NM, SFT Initialization=true2025.0326.15023.11018.724.323.71.4
GFlowVLM w/ Var-TBAssump.=NM, SFT Initialization=false2025.0325.75022.210.216.122.721.91.41
RL4VLMAssump.=NM, SFT Initialization=true2025.0322.149.113.59.815.220.120.61.11
RL4VLMAssump.=M, SFT Initialization=true2025.0321.747.414.710.414.418.8181.12
ProCeedRLBackbone=Qwen3-1.7B2026.0420.35———————
DAPOBackbone=Qwen3-1.7B2026.0418.69———————
SFT-w/o-Assump.=-2025.0317.139.2014.411.1028.61.06
SFT-w/-Assump.=-2025.0315.932.7010.310.8021.81.02
Qwen3-1.7BBackbone=Qwen3-1.7B2026.0411.07———————