List your company

Theory of Mind on BigToM

99.5AccuracyRECTOM
15.2637.135980.87May 19, 2026May 22, 2026May 26, 2026May 30, 2026Jun 2, 2026Jun 6, 2026Jun 10, 2026

Evaluation results

MethodLinks
RECTOMModel=GPT-5.42026.0699.5
CoTModel=GPT-5.42026.0699
SIMTOMModel=GPT-5.42026.0699
RECTOMModel=Gemini-32026.0699
RECTOMModel=Gemma-42026.0699
CoTModel=Gemini-32026.0698.75
UserHarnessBackbone=Claude-Opus-4.72026.0598.67
RECTOMModel=Qwen3.52026.0698.5
Ledger promptingBackbone=GPT-5.42026.0598.25
UserHarnessBackbone=Gemini-3.1-Pro2026.0598.25
UserHarnessBackbone=Claude-Sonnet-4.62026.0598.17
SIMTOMModel=Gemini-32026.0698
TIMETOMModel=GPT-5.42026.0697.75
UserHarnessBackbone=GPT-5.42026.0597.67
Ledger promptingBackbone=Qwen3-14B2026.0597.25
Ledger promptingBackbone=GPT-OSS-120B2026.0597.08
Ledger promptingBackbone=GPT-5.4-mini2026.0596.67
Ledger promptingBackbone=Qwen3-32B2026.0596.5
CoTModel=Qwen3.52026.0696.25
Ledger promptingBackbone=GPT-OSS-20B2026.0596.17
UserHarnessBackbone=Qwen3-32B2026.0595.42
UserHarnessBackbone=GPT-5.4-mini2026.0595.25
SIMTOMModel=Qwen3.52026.0694.75
UserHarnessBackbone=Qwen3-14B2026.0594.17
Paradox promptingBackbone=GPT-5.42026.0593.83
UserHarnessBackbone=GPT-OSS-120B2026.0593.83
Ledger promptingBackbone=Qwen3-8B2026.0593.5
CoTModel=Gemma-42026.0693.25
TIMETOMModel=Gemini-32026.0692.5
Ledger promptingBackbone=Llama-3.1-8B2026.0592.17
Direct promptingBackbone=GPT-OSS-120B2026.0592
Paradox promptingBackbone=GPT-OSS-120B2026.0591.83
TIMETOMModel=Qwen3.52026.0691
Mistral-NeMo-12BParams=12B, Inference Protocol=direct model inference2026.0590.5
UserHarnessBackbone=GPT-OSS-20B2026.0590.17
OSCToM-8BParams=8B, Inference Protocol=direct model inference2026.0589.8
UserHarnessBackbone=Qwen3-8B2026.0589.67
Direct promptingBackbone=Qwen3-32B2026.0588.5
SIMTOMModel=Gemma-42026.0688
Paradox promptingBackbone=GPT-OSS-20B2026.0587.83
Paradox promptingBackbone=Qwen3-32B2026.0587.33
Direct promptingBackbone=GPT-5.42026.0586.92
AutoToMBackbone=GPT-4o2026.0586.92
Direct promptingBackbone=GPT-OSS-20B2026.0586.75
UserHarnessBackbone=Gemini-3-Flash2026.0586.25
Llama-3.1-8B-BaseParams=8B, Inference Protocol=direct model inference2026.0586
Phi-3-Medium-14BParams=14B, Inference Protocol=direct model inference2026.0585.5
UserHarnessBackbone=Llama-3.1-8B2026.0582.83
ExploreToMParams=8B, Inference Protocol=iterative A* heuristic search2026.0581
TIMETOMModel=Gemma-42026.0680
Paradox promptingBackbone=GPT-5.4-mini2026.0577.92
SimToMSetting=reported setting2026.0577.5
Paradox promptingBackbone=Qwen3-14B2026.0577.08
Direct promptingBackbone=GPT-5.4-mini2026.0576.42
Paradox promptingBackbone=Qwen3-8B2026.0576.42
Direct promptingBackbone=Llama-3.1-8B2026.0574.42
Paradox promptingBackbone=Llama-3.1-8B2026.0574
Direct promptingBackbone=Qwen3-14B2026.0573.5
Direct promptingBackbone=Qwen3-8B2026.0572.33
LIMPSetting=reported setting2026.0561.67
BIP-ALMSetting=reported setting2026.0550.33
Qwen2.5-32BParams=32B, Inference Protocol=direct model inference2026.0547.5
Qwen2.5-14BParams=14B, Inference Protocol=direct model inference2026.0529.5
Gemma-2-27BParams=27B, Inference Protocol=direct model inference2026.0518.5