Go-Explore: a New Approach for Hard-Exploration Problems
read the original abstract
A grand challenge in reinforcement learning is intelligent exploration, especially when rewards are sparse or deceptive. Two Atari games serve as benchmarks for such hard-exploration domains: Montezuma's Revenge and Pitfall. On both games, current RL algorithms perform poorly, even those with intrinsic motivation, which is the dominant method to improve performance on hard-exploration domains. To address this shortfall, we introduce a new algorithm called Go-Explore. It exploits the following principles: (1) remember previously visited states, (2) first return to a promising state (without exploration), then explore from it, and (3) solve simulated environments through any available means (including by introducing determinism), then robustify via imitation learning. The combined effect of these principles is a dramatic performance improvement on hard-exploration problems. On Montezuma's Revenge, Go-Explore scores a mean of over 43k points, almost 4 times the previous state of the art. Go-Explore can also harness human-provided domain knowledge and, when augmented with it, scores a mean of over 650k points on Montezuma's Revenge. Its max performance of nearly 18 million surpasses the human world record, meeting even the strictest definition of "superhuman" performance. On Pitfall, Go-Explore with domain knowledge is the first algorithm to score above zero. Its mean score of almost 60k points exceeds expert human performance. Because Go-Explore produces high-performing demonstrations automatically and cheaply, it also outperforms imitation learning work where humans provide solution demonstrations. Go-Explore opens up many new research directions into improving it and weaving its insights into current RL algorithms. It may also enable progress on previously unsolvable hard-exploration problems in many domains, especially those that harness a simulator during training (e.g. robotics).
This paper has not been read by Pith yet.
Forward citations
Cited by 20 Pith papers
-
Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution
Promptbreeder evolves both task prompts and the mutation prompts that improve them using LLMs, outperforming Chain-of-Thought and Plan-and-Solve on arithmetic and commonsense reasoning benchmarks.
-
Decision Transformer: Reinforcement Learning via Sequence Modeling
Decision Transformer casts RL as autoregressive sequence modeling conditioned on desired returns, past states and actions, matching or exceeding offline RL baselines on Atari, Gym and Key-to-Door tasks.
-
TriSearch: Learning to Optimize Triangulations via Bistellar Flips
TriSearch is an RL framework that optimizes triangulations of polytopes using bistellar flips with a circuit-supported subtriangulation action representation, generalizing zero-shot to larger instances and outperformi...
-
Baba in Wonderland: Online Self-Supervised Dynamics Discovery for Executable World Models
Alice uses preservation conflicts from failed candidate updates to create class-stratified hypotheses and guide exploration, improving executable world-model learning under prior misalignment.
-
Voyager: An Open-Ended Embodied Agent with Large Language Models
Voyager achieves superior lifelong learning in Minecraft by combining an automatic exploration curriculum, a library of executable skills, and iterative LLM prompting with environment feedback, yielding 3.3x more uniq...
-
The Red Queen G\"odel Machine: Co-Evolving Agents and Their Evaluators
The Red Queen Gödel Machine organizes recursive self-improvement into epochs with fixed intra-epoch evaluation while allowing utility evolution at boundaries, yielding reported gains on coding, paper writing, and proo...
-
Uncovering Linguistic Fragility in Vision-Language-Action Models via Diversity-Aware Red Teaming
DAERT generates diverse adversarial instructions via a uniform policy in RL to drop VLA task success rates from 93.33% to 5.85% on benchmarks with models like π0 and OpenVLA.
-
Cognitive Architectures for Language Agents
CoALA is a modular cognitive architecture for language agents that organizes memory components, action spaces for internal and external interaction, and a generalized decision-making loop to support more systematic de...
-
Learning World Graphs to Accelerate Hierarchical Reinforcement Learning
A two-stage framework learns a world graph of pivotal states task-agnostically via joint training of a latent model and curiosity-driven policy, then uses the graph to accelerate hierarchical RL on maze tasks.
-
Position: RL Researchers Need to Distinguish Between Solving Simulators and Using Simulators as a Proxy
RL researchers need to separate 'solving simulators' from 'using simulators as proxies' to prevent misleading conclusions about algorithm performance.
-
Mesh-RL: Coupled subgrid reinforcement learning
Mesh-RL applies finite-element-inspired domain decomposition with overlapping subgrids to accelerate temporal-difference learning across distant states in grid-world environments.
-
The Red Queen G\"odel Machine: Co-Evolving Agents and Their Evaluators
RQGM enables co-evolution of agents and evaluators across epochs with non-stationary utilities, reporting gains in coding pass rates, paper acceptance, and proof grading over prior self-improving agents.
-
ConTrack: Constrained Hand Motion Tracking with Adaptive Trade-off Control
ConTrack introduces a constrained RL method with online dual-variable adaptation and adaptive resets for improved long-horizon hand tracking in simulation and on real robots.
-
REFLECTOR: Internalizing Step-wise Reflection against Indirect Jailbreak
Reflector trains LLMs to internalize step-wise self-reflection through SFT on teacher data followed by RL with outcome and validity rewards, reporting over 90% defense success against indirect jailbreaks and a 5.85% g...
-
REFLECTOR: Internalizing Step-wise Reflection against Indirect Jailbreak
Reflector internalizes step-wise self-reflection in LLMs via teacher-guided SFT then RL with outcome and validity rewards, claiming over 90% defense success against indirect jailbreaks plus utility gains like 5.85% on GSM8K.
-
DyGRO-VLA: Cross-Task Scaling of Vision-Language-Action Models via Dynamic Grouped Residual Optimization
DyGRO-VLA is a two-stage optimization framework for cross-task scaling of Vision-Language-Action models via dynamic grouped residual optimization in RL.
-
Mind Dreamer: Untethering Imagination via Active Causal Intervention on Latent Manifolds
Mind Dreamer uses active causal intervention via an adversarial initial-state generator and relay value functions to untether imagination in MBRL, claiming 1.67x average and up to 8.8x sparse-reward speedups over DreamerV3.
-
Mind Dreamer: Untethering Imagination via Active Causal Intervention on Latent Manifolds
Mind Dreamer adds an adversarial generator for non-continuous latent starting states plus Relay Value and Uncertainty Functions to handle credit assignment across jumps, claiming 1.67x average and 8.8x sparse-reward s...
-
PokeRL: Reinforcement Learning for Pokemon Red
PokeRL trains PPO agents to finish early Pokemon Red tasks using a loop-aware environment wrapper, multi-layer anti-loop mechanisms, and dense hierarchical rewards.
-
Optimistic Proximal Policy Optimization
OPPO augments PPO with optimistic policy evaluation driven by return uncertainty estimates and shows improved results over prior methods on a tabular sparse-reward task.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.