List your company

Auction Game Playing on Auction

0.75Payoff per RoundSEPO
0.504560.568280.6320.69572May 29, 2026

Evaluation results

MethodLinks
SEPOBase Model=Gemma 4 E4B-it, Training Stage=SEPO (s75)2026.050.750.0210.208-0.90.13
SFTBase Model=Gemma 4 E4B-it, Training Stage=SFT2026.050.64600.208-1.0540.149
SEPOBase Model=Qwen 3.5-4B, Training Stage=SEPO (final)2026.050.5630.1250.255-1.20.122
SFTBase Model=Qwen 3.5-4B, Training Stage=SFT2026.050.5420.250.344-1.7830.23
BaseBase Model=Gemma 4 E4B-it, Training Stage=Base2026.050.5140.1460.38-2.0490.478
BaseBase Model=Qwen 3.5-4B, Training Stage=Base2026.050.5140.0420.49-2.8610.373