AI Paper Atlas

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

Mind the Budget: Accelerating Deep Reinforcement Learning using Constrained Early Exit Neural Networks

Julien Brosseit, Jasper Hoffmann, Joschka Boedecker

ICML 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Actor-critic theory

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
Add budgeted early-exit actor (BEXA) to SAC/TD3SAC and TD3 · MuJoCo 5 tasksActor FLOPs speedup: up to +584% FLOPs, +95% wall-clockImproved
BEXA with budget on harder tasks such as HumanoidBEXA · MuJoCo Humanoid-v4Actor FLOPs speedup: +9% (SAC), +39% (TD3)Mixed
Use all-exit ensemble inference instead of early-exit inference (ablation)BEXA · MuJoCo small actorsMean return: 62.8 -> 26.8 (SAC)Hurt
Train early exits with imitation of final exit instead of DRL loss (ablation)BEXA · MuJoCo small actorsMean return: 62.8 -> 29.2 (SAC); 38.7 -> 34.0 (TD3)Hurt
Use advantage-over-backbone gating instead of LP budget gating (ablation)BEXA · MuJoCo small actorsMean return: 62.8 -> 58.3 (SAC); 38.7 -> 40.5 (TD3)Mixed
Use softmax-over-Q gating instead of LP budget gating (ablation)BEXA · MuJoCo small actorsMean return: 62.8 -> 50.3 (SAC); 38.7 -> 49.3 (TD3)Mixed
Use stepwise exit training instead of joint training (ablation)BEXA · MuJoCo small actorsMean return: 62.8 -> 27.8 (SAC); 38.7 -> 16.5 (TD3)Hurt
Greedy exit selection by Q-value without budget constraint (ablation)early-exit actor · MuJoCo SAC trainingAverage exit takenHurt
Use separate critic head per exit (ablation)BEXA · MuJoColearning stabilityImproved

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