METTA: Memory-Enhanced Test-Time Adaptation for Meta-Reinforcement Learning

METTA: Memory-Enhanced Test-Time Adaptation for Meta-Reinforcement Learning. It centres on Meta, and also names Inference. Reported by arXiv. Bharat Hunt files it under AI Models — the section covering a new or updated model, its capabilities, benchmarks or availability.

Written by Bharat Hunt from the headline and the coverage below. The original reporting is the source of truth.