AI Paper Atlas

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

Safe Reinforcement Learning via Curriculum Induction

Matteo Turchetta, Andrey Kolobov, Shital Shah, Andreas Krause, Alekh Agarwal

NeurIPS 2020 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Intrinsically motivated exploration · 41 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 a curriculum optimized across students instead of no interventionCISR on Frozen Lake · Frozen LakeSuccess: 0.768±0.028 -> 0.960±0.004Improved
Use a curriculum optimized across students instead of no interventionCISR on Frozen Lake · Frozen LakeTraining failures: 3075.6±492.1 -> 0±0Improved
Use an optimized Narrow-to-Wide curriculum instead of no interventionCISR on Lunar Lander · Lunar Lander, noiseless 2-layer studentTraining failures: 1251.0±33.72 -> 1.7±0.13Improved
Use an optimized Narrow-to-Wide curriculum instead of no interventionCISR on Lunar Lander · Lunar Lander, noiseless 2-layer studentReturn: 225.1±7.4 -> 236.1±0.9No clear effect
Use an optimized curriculum instead of no interventionCISR on Lunar Lander · Lunar Lander, noisy 2-layer studentTraining failures: 1695.8±31.04 -> 2.6±0.23Improved
Transfer the optimized curriculum to a 1-layer studentCISR on Lunar Lander · Lunar Lander, noiseless 1-layer studentTraining failures: 1175.0±80.56 -> 1.9±0.13Improved
Use Narrow rather than the optimized Narrow-to-Wide curriculum (ablation)CISR on Lunar Lander · Lunar Lander, noiseless 2-layer studentReturn: 187.2±9.0 vs 236.1±0.9Hurt
Use Wide rather than the optimized Narrow-to-Wide curriculum (ablation)CISR on Lunar Lander · Lunar Lander, noiseless 2-layer studentReturn: 199.5±8.3 vs 236.1±0.9Hurt
Use HR rather than an optimized curriculum (ablation)CISR on Frozen Lake · Frozen LakeReturn: -0.222±0.013 vs 5.368±0.025Hurt
Use SR1 rather than an optimized curriculum (ablation)CISR on Frozen Lake · Frozen LakeReturn: 4.839±0.065 vs 5.368±0.025Hurt

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