List your company

Planning on 8x8 two-room gridworld (test)

0.89Validity (%)L-ICL
0.13080.32790.5250.7221Jan 30, 2026

Evaluation results

MethodLinks
L-ICLBase Model=DeepSeek V3, Number of examples (m)=60, Character count=2k chars2026.010.890.890.77
Self-ConsistencyBase Model=DeepSeek V3, k value=52026.010.590.450.43
Self-RefineBase Model=DeepSeek V3, k value=52026.010.510.440.38
ReActBase Model=DeepSeek V3, Prompting Strategy=ReAct2026.010.480.410.37
PTPBase Model=DeepSeek V3, m value=02026.010.40.330.28
RAG-ICLBase Model=DeepSeek V3, Character count=20k chars2026.010.210.090.09
RAG-ICLBase Model=DeepSeek V3, Character count=10k chars2026.010.20.060.06
Zero-ShotBase Model=DeepSeek V3, Prompting Strategy=Zero-Shot2026.010.1600