Hessel et al.: Rainbow integration and ablation of DQN improvements

What changed

Integrated six complementary DQN extensions and used ablations to characterize their joint effects on Atari learning efficiency and final performance.

Scientific scope

Integrated six complementary DQN extensions and used ablations to characterize their joint effects on Atari learning efficiency and final performance. Double Q, prioritized replay, dueling networks, multistep returns, distributional targets and noisy exploration retain original credit. This is integration/interaction evidence, not six new inventions. Preprint2017 and AAAI publication2018 are distinct dates.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Double Q, prioritized replay, dueling networks, multistep returns, distributional targets and noisy exploration retain original credit. This is integration/interaction evidence, not six new inventions. Preprint2017 and AAAI publication2018 are distinct dates.

Contributors

• Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.

• Rainbow: Combining Improvements in Deep Reinforcement Learning

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction