AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

The Nature of Temporal Difference Errors in Multi-step Distributional Reinforcement Learning

Yunhao Tang, Remi Munos, Mark Rowland, Bernardo Avila Pires, Will Dabney, Marc G Bellemare

NeurIPS 2022 · Reinforcement learning › Policy gradient and actor-critic · approach: Distributional value learning · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
increase trace cutoff c-bar in distributional Retrace (ablation)Distributional Retrace (quantile) · Tabular MDPsfixed point quality L_pImproved
use uncorrected n-step instead of Retrace (ablation)multi-step distributional RL · Tabular MDPsfixed point biasHurt
multi-step Retrace (n=3) instead of one-stepC51 · Atari-57median human-normalized scoreImproved
Retrace instead of uncorrected n-stepC51 · Atari-57median human-normalized scoreImproved
multi-step variants (Retrace, n-step) instead of one-stepQR-DQN · Atari-57mean and median human-normalized scoreImproved
Retrace instead of uncorrected n-stepQR-DQN · Atari-57mean human-normalized scoreImproved
Retrace instead of uncorrected n-stepQR-DQN · Atari-57median human-normalized scoreNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy