Research map / Reinforcement learning / Visual and representation RL
One Solution is Not All You Need: Few-Shot Extrapolation via Structured MaxEnt RL
Saurabh Kumar, Aviral Kumar, Sergey Levine, Chelsea Finn
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 changed | Where | Measured | Effect |
|---|---|---|---|
| SMERL (diverse near-optimal policies with test-time selection) instead of SAC with 1 policy | SAC · HalfCheetah-Goal, Walker-Velocity, Hopper-Velocity with obstacle, force, motor failure | return | Improved |
| Use 5 latent-conditioned policies instead of 1 for SAC (ablation) | SAC · Perturbed test environments (except HalfCheetah-Goal + Obstacle) | return | Improved |
| Add diversity reward only when return is near-optimal (SMERL) instead of naively summing DIAYN and task rewards (SAC+DIAYN) (ablation) | SAC+DIAYN · Perturbed test environments | return | Mixed |
| Add task reward to DIAYN (SMERL instead of DIAYN alone) (ablation) | DIAYN · Perturbed test environments | return | Improved |
| Use SMERL instead of adversarial RARL (single policy) | RARL · Perturbed test environments | return | Improved |
| Choose best of 5 SMERL policies at test time versus the best training-time policy (ablation) | SMERL · HalfCheetah-Goal + Force, magnitude 300.0 | return: -424.1 (policy 5) -> -222.7 (policy 1) | Improved |
Closest papers by meaning
- Enabling Adaptive Agent Training in Open-Ended Simulators by Targeting Diversity (NeurIPS 2024 · similarity 0.89)
- DGPO: Discovering Multiple Strategies with Diversity-Guided Policy Optimization (AAAI 2024 · similarity 0.88)
- Skill-aware Mutual Information Optimisation for Zero-shot Generalisation in Reinforcement Learning (NeurIPS 2024 · similarity 0.87)
- Controlling Behavioral Diversity in Multi-Agent Reinforcement Learning (ICML 2024 · similarity 0.86)
- Generalizable Task Representation Learning for Offline Meta-Reinforcement Learning with Data Limitations (AAAI 2024 · similarity 0.86)
- Learning Policy Committees for Effective Personalization in MDPs with Diverse Tasks (ICML 2025 · similarity 0.86)
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.
