List your company

Trajectory QA on AMA-Bench Full 208-episode trajectory

38.2F1 ScoreVISTA
32.68834.11935.5536.981Jun 29, 2026

Evaluation results

MethodLinks
VISTARuntime/ep=43.7s, Tokens/ep=148.49K2026.0638.273.1
AMARuntime/ep=176.5s, Tokens/ep=268.98K2026.0636.875.3
EMem-styleRuntime/ep=166.2s, Tokens/ep=470.30K2026.0636.365.1
BM25Runtime/ep=35.5s, Tokens/ep=303.43K2026.0633.557.5
Mem0-styleRuntime/ep=108.0s, Tokens/ep=30.63K2026.0632.953.6