List your company

Multi-turn Agentic Task on ScienceWorld

62Success RateReMem
7.9221.963650.04Nov 25, 2025Dec 23, 2025Jan 20, 2026Feb 18, 2026Mar 18, 2026Apr 15, 2026May 14, 2026

Evaluation results

MethodLinks
ReMemLLM Backbone=Claude 3.7 Sonnet2025.116289
ReMemLLM Backbone=Gemini 2.5 Flash2025.115881
EVOLIB2026.0557.484.3
DCType=test-time learning (TTL)2026.0555.983.6
BaseType=base sampling2026.0553.382.9
ExpRecentLLM Backbone=Gemini 2.5 Flash2025.115383
ExpRAGLLM Backbone=Gemini 2.5 Flash2025.115381
ExpRAGType=test-time learning (TTL)2026.0551.180.4
ExpRecentLLM Backbone=Claude 3.7 Sonnet2025.114982
ExpRAGLLM Backbone=Claude 3.7 Sonnet2025.114676
ReActLLM Backbone=Claude 3.7 Sonnet2025.114477
Mem0LLM Backbone=Claude 3.7 Sonnet2025.113776
AmemLLM Backbone=Gemini 2.5 Flash2025.113674
ReActLLM Backbone=Gemini 2.5 Flash2025.113471
SelfRAGLLM Backbone=Gemini 2.5 Flash2025.113474
DC-RSLLM Backbone=Claude 3.7 Sonnet2025.113474
AWMLLM Backbone=Claude 3.7 Sonnet2025.113474
DC-RSLLM Backbone=Gemini 2.5 Flash2025.113373
AWMLLM Backbone=Gemini 2.5 Flash2025.113373
AmemLLM Backbone=Claude 3.7 Sonnet2025.113373
DC-CuLLM Backbone=Claude 3.7 Sonnet2025.113375
Mem0LLM Backbone=Gemini 2.5 Flash2025.113270
HistoryLLM Backbone=Claude 3.7 Sonnet2025.113274
HistoryLLM Backbone=Gemini 2.5 Flash2025.113171
SelfRAGLLM Backbone=Claude 3.7 Sonnet2025.113174
DC-CuLLM Backbone=Gemini 2.5 Flash2025.112971
BaselineLLM Backbone=Gemini 2.5 Flash2025.112459
BaselineLLM Backbone=Claude 3.7 Sonnet2025.111053