Research map / Reinforcement learning / Visual and representation RL
Diversity-Driven Exploration Strategy for Deep Reinforcement Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Add a policy-distance exploration term | DQN · Deceptive Reward 50×50 | average reward: 0.010 -> 0.202; 0.009 -> 0.202 | Helped |
| Add a policy-distance exploration term | DQN · Deceptive Reward 100×100 | average reward: 0.010 -> 0.604; 0.004 -> 0.604 | Helped |
| Add a policy-distance exploration term | DQN · Deceptive Reward 200×200 | average reward: 0.010 -> 0.208; 0.010 -> 0.208 | Helped |
| Add a policy-distance exploration term | DQN · Sparse Reward 50×50 | average reward: 0.300 -> 1.000; 0.400 -> 1.000 | Helped |
| Add a policy-distance exploration term | DQN · Sparse Reward 100×100 | average reward: 0.100 -> 1.000; 0.200 -> 1.000 | Helped |
| Add a policy-distance exploration term | DQN · Sparse Reward 200×200 | average reward: 0.000 -> 1.000; 0.000 -> 1.000 | Helped |
| Use distance-based scaling instead of linear decay for the diversity term (ablation) | Div-DQN · Atari 2600 | median score | Mixed |
| Use proactive performance-based scaling instead of reactive scaling (ablation) | Div-A2C · Atari 2600 | median score | Mixed |
| Use distance-based scaling instead of linear decay for the diversity term (ablation) | Div-DDPG · MuJoCo | median score | Mixed |
| Use adaptive scaling for the diversity term (ablation) | Diversity-driven exploration · Benchmark environments | overall performance | Helped |
| Apply diversity-driven exploration to DQN | DQN · Montezuma's Revenge | score | No clear effect |
Closest papers by meaning
- Effective Diversity in Population Based Reinforcement Learning (NeurIPS 2020 · similarity 0.92)
- Improving Exploration in Evolution Strategies for Deep Reinforcement Learning via a Population of Novelty-Seeking Agents (NeurIPS 2018 · similarity 0.92)
- DGPO: Discovering Multiple Strategies with Diversity-Guided Policy Optimization (AAAI 2024 · similarity 0.89)
- Enhancing Diversity In Parallel Agents: A Maximum State Entropy Exploration Story (ICML 2025 · similarity 0.88)
- Rethinking Exploration in Reinforcement Learning with Effective Metric-Based Exploration Bonus (NeurIPS 2024 · similarity 0.88)
- The Curse of Diversity in Ensemble-Based Exploration (ICLR 2024 · similarity 0.88)
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.