ICLR 2021 accepted papers on Reinforcement learning
94 of the 860 papers accepted at ICLR 2021 (10.9%) are on Reinforcement learning, compared with 11.8% at ICLR 2020. They fall into 5 topics.
See ICLR 2021 in the interactive map
Submitting to ICLR? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Visual and representation RL · 41 papers
- Self-Supervised Policy Adaptation during Deployment (60 citations)
- Learning with AMIGo: Adversarially Motivated Intrinsic Goals (50 citations)
- Data-Efficient Reinforcement Learning with Self-Predictive Representations (47 citations)
- Parrot: Data-Driven Behavioral Priors for Reinforcement Learning (38 citations)
- On the role of planning in model-based deep reinforcement learning (35 citations)
- Modelling Hierarchical Structure between Dialogue Policy and Natural Language Generator with Option Framework for Task-oriented Dialogue System (31 citations)
- CausalWorld: A Robotic Manipulation Benchmark for Causal Structure and Transfer Learning (30 citations)
- Contrastive Behavioral Similarity Embeddings for Generalization in Reinforcement Learning (26 citations)
- Hierarchical Reinforcement Learning by Discovering Intrinsic Options (24 citations)
- Model-Based Visual Planning with Self-Supervised Functional Distances (23 citations)
- Return-Based Contrastive Representation Learning for Reinforcement Learning (21 citations)
- Evolving Reinforcement Learning Algorithms (19 citations)
and 29 more in the interactive map.
Policy gradient and actor-critic · 29 papers
- Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels (126 citations)
- What Matters for On-Policy Deep Actor-Critic Methods? A Large-Scale Study (47 citations)
- Robust Reinforcement Learning on State Observations with Learned Optimal Adversary (44 citations)
- Conservative Safety Critics for Exploration (39 citations)
- The Importance of Pessimism in Fixed-Dataset Policy Optimization (23 citations)
- Single-Timescale Actor-Critic Provably Finds Globally Optimal Policy (17 citations)
- Winning the L2RPN Challenge: Power Grid Management via Semi-Markov Afterstate Actor-Critic (16 citations)
- Quantifying Differences in Reward Functions (15 citations)
- Adversarially Guided Actor-Critic (14 citations)
- Genetic Soft Updates for Policy Evolution in Deep Reinforcement Learning (12 citations)
- Enforcing robust control guarantees within neural network policies (11 citations)
- Regularization Matters in Policy Optimization - An Empirical Study on Continuous Control (8 citations)
and 17 more in the interactive map.
Offline RL · 10 papers
- Deployment-Efficient Reinforcement Learning via Model-Based Offline Optimization (50 citations)
- What are the Statistical Limits of Offline RL with Linear Function Approximation? (35 citations)
- Risk-Averse Offline Reinforcement Learning (23 citations)
- Benchmarks for Deep Off-Policy Evaluation (21 citations)
- Model-Based Offline Planning (18 citations)
- OPAL: Offline Primitive Discovery for Accelerating Offline Reinforcement Learning (15 citations)
- Autoregressive Dynamics Models for Offline Policy Evaluation and Optimization (8 citations)
- FOCAL: Efficient Fully-Offline Meta-Reinforcement Learning via Distance Metric Learning and Behavior Regularization (8 citations)
- Representation Balancing Offline Model-based Reinforcement Learning (8 citations)
- DeepAveragers: Offline Reinforcement Learning By Solving Derived Non-Parametric MDPs (4 citations)
Imitation learning · 8 papers
- Primal Wasserstein Imitation Learning (36 citations)
- Learning to Reach Goals via Iterated Supervised Learning (23 citations)
- Scalable Bayesian Inverse Reinforcement Learning (13 citations)
- Behavioral Cloning from Noisy Demonstrations (10 citations)
- Solving Compositional Reinforcement Learning Problems via Task Reduction (9 citations)
- Learning "What-if" Explanations for Sequential Decision-Making (5 citations)
- Text Generation by Learning from Demonstrations (5 citations)
- Regularized Inverse Reinforcement Learning (2 citations)
MDPs and planning · 6 papers
- Vulnerability-Aware Poisoning Mechanism for Online RL with Unknown Dynamics (14 citations)
- Greedy-GQ with Variance Reduction: Finite-time Analysis and Improved Complexity (7 citations)
- Learning Robust State Abstractions for Hidden-Parameter Block MDPs (5 citations)
- Acting in Delayed Environments with Non-Stationary Markov Policies (4 citations)
- Provable Rich Observation Reinforcement Learning with Combinatorial Latent States (3 citations)
- Efficient Reinforcement Learning in Factored MDPs with Application to Constrained RL (2 citations)
