English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Offline reinforcement learning (RL) aims at learning a good policy from abatch of collected data, without extra interactions with the environment duringtraining. However, current offline RL benchmarks commonly have a large realitygap, because they involve large datasets collected by highly exploratorypolicies, and the trained policy is directly evaluated in the environment. Inreal-world situations, running a highly exploratory policy is prohibited toensure system safety, the data is commonly very limited, and a trained policyshould be well validated before deployment. In this paper, we present a nearreal-world offline RL benchmark, named NeoRL, which contains datasets fromvarious domains with controlled sizes, and extra test datasets for policyvalidation. We evaluate existing offline RL algorithms on NeoRL and argue thatthe performance of a policy should also be compared with the deterministicversion of the behavior policy, instead of the dataset reward. The empiricalresults demonstrate that the tested offline RL algorithms become lesscompetitive to the deterministic policy on many datasets, and the offlinepolicy evaluation hardly helps. The NeoRL suit can be found athttp://polixir.ai/research/neorl. We hope this work will shed some light onfuture research and draw more attention when deploying RL in real-worldsystems.

NeoRL: A Near Real-World Benchmark for Offline Reinforcement Learning