AI Paper Atlas

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

VIREL: A Variational Inference Framework for Reinforcement Learning

Matthew Fellows, Anuj Mahajan, Tim G. J. Rudner, Shimon Whiteson

NeurIPS 2019 · Reinforcement learning › Policy gradient and actor-critic · approach: Off-policy actor-critic with expressive policies · 29 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 VIREL or beta instead of SACSAC · Humanoid-v2 and Ant-v2episodic returnImproved
Use VIREL or beta instead of DDPGDDPG · Humanoid-v2 and Ant-v2episodic returnImproved
Use VIREL or beta instead of SAC and DDPGSAC and DDPG · Simple low-dimensional MuJoCo tasksepisodic returnNo clear effect
Use VIREL or beta instead of SACSAC · Half-Cheetahepisodic returnNo clear effect
Use VIREL1 or VIREL2 instead of SACSAC with two soft critics · Humanoid-gym and Humanoid-rllabepisodic returnImproved
Use an estimated adaptive entropy scale instead of a constant entropy scale (ablation)VIREL actor update · High-dimensional MuJoCo tasksepisodic returnNo clear effect
Use two hard Q-functions instead of the standard VIREL critic setup (ablation)VIREL1 · Half-Cheetah and humanoid tasksepisodic returnMixed
Use one hard and one soft Q-function (ablation)VIREL2 · Half-Cheetah and humanoid tasksepisodic returnMixed
Remove the scalar lambda that scales the estimated residual error (ablation)beta · MuJoCo continuous-control tasksperformance stabilityHurt

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