Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Long-horizon memory Question Answering research benchmarks · SOTA2 Research

Long-horizon memory Question Answering

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

GPT-4o (ceiling)

73.7Accuracy

9

May 6, 2026

GPT-4o (ceiling)

68.4Accuracy

9

May 6, 2026

ATM-Bench subset431

69.61EM

3

May 29, 2026

Popular datasets

LongBench LongMemEval ATM-Bench

Follow for updates

Join SOTA2 Contact