List your company
ResearchFormerly WizWand

Causal Reasoning on BBH Causal Judgement

78Accuracy (BBH Causal Judgement)evaluation-instructed prompt optimization
5258.7565.572.25Nov 25, 2025Dec 24, 2025Jan 22, 2026Feb 20, 2026Mar 21, 2026Apr 19, 2026May 18, 2026
Updated 5mo ago

Evaluation results

MethodLinks
evaluation-instructed prompt optimizationBackbone=GPT-4o, Optimization Type=Query-dependent2025.1178
Pro-RefineBackbone=GPT-4o, Optimization Type=Query-dependent2025.1175
Self-RefineBackbone=GPT-4o, Optimization Type=Query-dependent2025.1174
TextGradBackbone=GPT-4o, Optimization Type=Template2025.1174
LLM onlyBackbone=GPT-4o, Optimization Type=None2025.1173
PSAO_LLMBackbone Model=Gemini-2-Flash, Prompt Optimization Technique=PSAO_LLM2026.0572.73
APEBackbone=GPT-4o, Optimization Type=Template2025.1172
MIPROModel=GPT-4o2026.0570.91
MIPROLLM=GPT-4o2026.0570.91
GEPABackbone Model=Gemini-2-Flash, Prompt Optimization Technique=GEPA2026.0570.7
COPROModel=GPT-4o2026.0570
COPROLLM=GPT-4o2026.0570
PSAO_LLMModel=GPT-4o2026.0569.07
PSAO_LLMLLM=GPT-4o2026.0569.07
ProTeGiBackbone Model=Gemini-2-Flash, Prompt Optimization Technique=ProTeGi2026.0568.91
COPROBackbone Model=Gemini-2-Flash, Prompt Optimization Technique=COPRO2026.0568.91
BaselineLLM=GPT-4o2026.0568.73
PromptAgentBackbone Model=Gemini-2-Flash, Prompt Optimization Technique=PromptAgent2026.0568.55
GEPALLM=GPT-4o2026.0568
ProTeGiLLM=GPT-4o2026.0567.27
MIPROBackbone Model=Gemini-2-Flash, Prompt Optimization Technique=MIPRO2026.0567.09
BaselineBackbone Model=Gemini-2-Flash, Prompt Optimization Technique=Baseline2026.0566.36
PromptAgentLLM=GPT-4o2026.0565.09
evaluation-instructed prompt optimizationBackbone=LLaMA-3.1, Optimization Type=Query-dependent2025.1164
evaluation-instructed prompt optimizationBackbone=LLaMA-3, Optimization Type=Query-dependent2025.1163
TextGradBackbone=LLaMA-3, Optimization Type=Template2025.1162
Pro-RefineBackbone=LLaMA-3, Optimization Type=Query-dependent2025.1161
Pro-RefineBackbone=LLaMA-3.1, Optimization Type=Query-dependent2025.1161
LLM onlyBackbone=LLaMA-3, Optimization Type=None2025.1160
APEBackbone=LLaMA-3, Optimization Type=Template2025.1159
Self-RefineBackbone=LLaMA-3.1, Optimization Type=Query-dependent2025.1159
ReElicitevaluations=30 prompt evaluations2026.0558.4
Self-RefineBackbone=LLaMA-3, Optimization Type=Query-dependent2025.1158
LLM onlyBackbone=LLaMA-3.1, Optimization Type=None2025.1158
OPROevaluations=30 prompt evaluations2026.0557.1
APEevaluations=30 prompt evaluations2026.0557
TextGradBackbone=LLaMA-3.1, Optimization Type=Template2025.1157
TextGradevaluations=30 prompt evaluations2026.0556
PromptBreederevaluations=30 prompt evaluations2026.0555.8
APEBackbone=LLaMA-3.1, Optimization Type=Template2025.1153