List your company

Interactive Decision Making on ScienceWorld (OOD)

9.9ScoreMTRouter
-28.372-18.436-8.51.436Apr 26, 2026

Evaluation results

MethodLinks
MTRouterRouting Level=Multi-Turn Routers (turn-level)2026.049.916.3
RouterDCRouting Level=Single-Turn Routers (episode-level)2026.045.52.5
EmbedLLMRouting Level=Single-Turn Routers (episode-level)2026.0453
GPT-5Routing Level=Single-Model Baselines2026.044.947.6
AvengersProRouting Level=Single-Turn Routers (episode-level)2026.042.44.1
Router-R1Routing Level=Multi-Turn Routers (turn-level)2026.042.121
GPT-OSS-120BRouting Level=Single-Model Baselines2026.041.14.2
MiniMax-M2Routing Level=Single-Model Baselines2026.040.910.9
Kimi-K2Routing Level=Single-Model Baselines2026.040.28.9
LLM RouterRouting Level=Multi-Turn Routers (turn-level)2026.04-0.428.3
Gemini-2.5-Flash-LiteRouting Level=Single-Model Baselines2026.04-2.11.5
DeepSeek-V3.2Routing Level=Single-Model Baselines2026.04-4.222.8
Random RouterRouting Level=Multi-Turn Routers (turn-level)2026.04-8.120.3
OpenRouterRouting Level=Multi-Turn Routers (turn-level)2026.04-26.915.5