Reproduce “Permanent and Transient Representations for Continual Reinforcement Learning” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Nishanth Anand, Doina Precup

The authors develop a continual reinforcement learning method that combines a slowly learned permanent value estimate with a transient correction, using separate representations and a MinHash-based memory. In Craftax-Classic, it reaches return 10 before 50 million environment steps and return 15 within 150 million, sooner than PQN(32); this demonstrates gains in environment steps, not wall-clock time, on the benchmark.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.