Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Meta-trained agents implement Bayes-optimal agents
Vladimir Mikulik, Grégoire Delétang, Tom McGrath, Tim Genewein, Miljan Martic, Shane Legg, Pedro Ortega
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 |
|---|---|---|---|
| Train a memory-based RNN meta-learner on the task distribution | RNN meta-learner · 14 prediction and bandit tasks | behavioral dissimilarity from Bayes-optimal agent | Improved |
| Train a memory-based RNN meta-learner | RNN meta-learner · Prediction and bandit tasks | behavioral distance from Bayes-optimal policy | Improved |
| Use the Bayes-optimal agent to simulate the trained RNN | Trained RNN meta-learner · Prediction and bandit tasks | output dissimilarity | Improved |
| Use the trained RNN to simulate the Bayes-optimal agent | Bayes-optimal agent · Prediction and bandit tasks | output dissimilarity | Mixed |
| Use an untrained RNN instead of a trained RNN (ablation) | RNN meta-learner · Prediction and bandit tasks | structural similarity to Bayes-optimal agent | Hurt |
| Replace recurrent memory with a feedforward context window (ablation) | Reduced-memory baseline · Prediction and bandit tasks | performance | Hurt |
| Increase the feedforward context window to the episode length (ablation) | Reduced-memory baseline · Dirichlet and high-precision Gaussian prediction tasks | performance | Improved |
| Increase the feedforward context window (ablation) | Reduced-memory baseline · Some prediction and bandit tasks | performance | Mixed |
Closest papers by meaning
- Been There, Done That: Meta-Learning with Episodic Recall (ICML 2018 · similarity 0.89)
- Meta-Learning for Simple Regret Minimization (AAAI 2023 · similarity 0.89)
- Meta-Cognitive Memory Policy Optimization for Long-Horizon LLM Agents (NeurIPS 2026 · similarity 0.88)
- Learning to Continually Learn via Meta-learning Agentic Memory Designs (NeurIPS 2026 · similarity 0.88)
- Predictive Coding Enhances Meta-RL To Achieve Interpretable Bayes-Optimal Belief Representation Under Partial Observability (NeurIPS 2025 · similarity 0.88)
- Test-Time Regret Minimization in Meta Reinforcement Learning (ICML 2024 · similarity 0.87)
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.
