List your company

Theory of Mind on ToMi

100AccuracyUserHarness
1.7227.23552.7578.265May 19, 2026May 20, 2026May 21, 2026May 22, 2026May 23, 2026May 24, 2026May 26, 2026

Evaluation results

MethodLinks
UserHarnessBackbone=Qwen3-8B2026.05100
UserHarnessBackbone=Qwen3-14B2026.05100
UserHarnessBackbone=Qwen3-32B2026.05100
UserHarnessBackbone=Llama-3.1-8B2026.05100
UserHarnessBackbone=GPT-OSS-20B2026.05100
UserHarnessBackbone=GPT-OSS-120B2026.05100
UserHarnessBackbone=GPT-5.4-mini2026.05100
UserHarnessBackbone=GPT-5.42026.05100
UserHarnessBackbone=Gemini-3-Flash2026.05100
UserHarnessBackbone=Gemini-3.1-Pro2026.05100
UserHarnessBackbone=Claude-Sonnet-4.62026.05100
UserHarnessBackbone=Claude-Opus-4.72026.05100
SymbolicToMSetting=reported setting2026.0598.6
ExploreToMParams=8B, Inference Protocol=iterative A* heuristic search2026.0595
DITTOBackbone=Qwen3-VL-8B-Instruct2026.0593
Paradox promptingBackbone=GPT-5.42026.0589.1
AutoToMBackbone=GPT-4o2026.0588.3
GPT-5.42026.0588
Ledger promptingBackbone=GPT-5.42026.0585.72
GPT-5-nano2026.0585
Paradox promptingBackbone=GPT-5.4-mini2026.0583.97
GRPOBackbone=Qwen3-VL-8B-Instruct2026.0582
Direct promptingBackbone=GPT-OSS-120B2026.0580.22
SimToMSetting=reported setting2026.0579.9
Direct promptingBackbone=GPT-5.42026.0579.72
OSCToM-8BParams=8B, Inference Protocol=direct model inference2026.0579.5
Direct promptingBackbone=GPT-5.4-mini2026.0578.36
Ledger promptingBackbone=GPT-OSS-120B2026.0577.38
Phi-3-Medium-14BParams=14B, Inference Protocol=direct model inference2026.0576
Paradox promptingBackbone=GPT-OSS-120B2026.0575.05
Qwen2.5-32BParams=32B, Inference Protocol=direct model inference2026.0574.7
Ledger promptingBackbone=GPT-5.4-mini2026.0573.03
HumanLM-8B2026.0572
Direct promptingBackbone=GPT-OSS-20B2026.0570.08
Paradox promptingBackbone=GPT-OSS-20B2026.0569.56
Qwen3-VL-8B-InstructRole=Base2026.0568
Mistral-NeMo-12BParams=12B, Inference Protocol=direct model inference2026.0566.8
Llama-3.1-8B-BaseParams=8B, Inference Protocol=direct model inference2026.0564.3
Ledger promptingBackbone=GPT-OSS-20B2026.0562.37
BIP-ALMSetting=reported setting2026.0555.6
Direct promptingBackbone=Qwen3-8B2026.0554.52
Qwen2.5-14BParams=14B, Inference Protocol=direct model inference2026.0553.8
Gemma-2-27BParams=27B, Inference Protocol=direct model inference2026.0553.3
Direct promptingBackbone=Llama-3.1-8B2026.0552.48
Paradox promptingBackbone=Qwen3-32B2026.0550.38
Paradox promptingBackbone=Qwen3-14B2026.0548.96
Ledger promptingBackbone=Qwen3-32B2026.0547.34
LIMPSetting=reported setting2026.0544.6
Ledger promptingBackbone=Qwen3-8B2026.0526.44
Ledger promptingBackbone=Qwen3-14B2026.0525.28
Paradox promptingBackbone=Qwen3-8B2026.0517.32
Direct promptingBackbone=Qwen3-32B2026.0514.02
Ledger promptingBackbone=Llama-3.1-8B2026.0511.24
Direct promptingBackbone=Qwen3-14B2026.056.2
Paradox promptingBackbone=Llama-3.1-8B2026.055.5