AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

Interval timing in deep reinforcement learning agents

Ben Deverett, Ryan Faulkner, Meire Fortunato, Gregory Wayne, Joel Z. Leibo

NeurIPS 2019 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Intrinsically motivated exploration · 10 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
use an LSTM controller rather than a feedforward controllerA3C-based interval reproduction agent · Interval reproduction taskaccuracyImproved
use an LSTM controller rather than a feedforward controllerA3C-based interval reproduction agent · Untrained sample intervalsgeneralizationImproved
vary the number of controller hidden units from 128 (ablation)LSTM and feedforward agents · Interval reproduction taskaccuracyNo clear effect
replace the LSTM with a vanilla RNN (ablation)Recurrent interval reproduction agent · Interval reproduction tasktask learningMixed
replace the LSTM with a GRU (ablation)Recurrent interval reproduction agent · Interval reproduction tasktask learningImproved
replace the LSTM with an RMC (ablation)Recurrent interval reproduction agent · Interval reproduction tasktask learningMixed
freeze initially randomized LSTM parameters (ablation)LSTM interval reproduction agent · Interval reproduction taskaccuracyNo clear effect
use 10-step chunks for policy and baseline updates instead of 100-step trajectories (ablation)LSTM interval reproduction agent · Interval reproduction taskaccuracyNo clear effect
use 10-step chunks for policy and baseline updates instead of 100-step trajectories (ablation)Feedforward interval reproduction agent · Interval reproduction taskaccuracyHurt

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