List your company

Sequential Decision-Making on Jericho ID (meta-train)

270.5Score (Detective)META-TTL
84.548132.824181.1229.376Apr 1, 2026

Evaluation results

MethodLinks
META-TTLArchitecture=GPT-5 (as Meta-Agent)2026.04270.553.78.1110.882162441
META-TTLArchitecture=GLM-5 (as Meta-Agent)2026.0422447.37.492.968142235
NaiveArchitecture=GLM-5 (as Meta-Agent)2026.04122.739.23.755.237111019
META-TTLArchitecture=Gemini 3 Flash (as Meta-Agent)2026.04115.6437.255.333132223
NaiveArchitecture=GPT-5 (as Meta-Agent)2026.04107.938.74.750.431111318
NaiveArchitecture=Gemini 3 Flash (as Meta-Agent)2026.04104.940.8449.929121117
Memory AgentArchitecture=Single-Agent Baselines2026.04103.742.7550.528131418
ReflexionArchitecture=Single-Agent Baselines2026.04100.741.44.348.828121217
StaticArchitecture=Single-Agent Baselines2026.0491.738.5545.124111416