Research map / Multi-agent systems and games / Human-AI and goal-conditioned
A Direct Approximation of AIXI Using Logical State Abstractions
Samuel Yang-Zhao, Tianyu Wang, Kee Siong Ng
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 |
|---|---|---|---|
| Use Phi-AIXI-CTW instead of U-Tree and PARSS-DT | RL agent · Biased RPS, Taxi, Jackpot, Stop Heist | learning curve reward | Improved |
| Use Phi-AIXI-CTW instead of U-Tree and PARSS | RL agent · Jackpot and Taxi | learning curve reward | Improved |
| Use Phi-AIXI-CTW instead of U-Tree | RL agent · Biased RPS and Stop Heist | learning curve reward | No clear effect |
| Use Phi-AIXI-CTW instead of U-Tree and PARSS | RL agent · Epidemic control, lambda=1 | learning curve reward | Improved |
| Use RF-BDD instead of Random Forest feature selection (ablation) | Phi-AIXI-CTW · Epidemic control | learning curve reward | Mixed |
| Increase vaccination efficacy (eta 2,4 -> 10,20) | Phi-AIXI-CTW with RF-BDD · Epidemic control, lambda=10 | average reward per cycle | Improved |
| Lower infection weighting lambda from 10 to 1 | Phi-AIXI-CTW · Epidemic control | agent activity | Mixed |
| Forward selection instead of Random Forest feature selection (ablation) | Phi-AIXI-CTW · Epidemic control | performance | No clear effect |
Closest papers by meaning
- Dynamic Knowledge Injection for AIXI Agents (AAAI 2024 · similarity 0.88)
- Bayesian Exploration Networks (ICML 2024 · similarity 0.86)
- Oracle-Efficient Reinforcement Learning for Max Value Ensembles (NeurIPS 2024 · similarity 0.85)
- Safeguarded Progress in Reinforcement Learning: Safe Bayesian Exploration for Control Policy Synthesis (AAAI 2024 · similarity 0.85)
- Autonomous Option Invention for Continual Hierarchical Reinforcement Learning and Planning (AAAI 2025 · similarity 0.84)
- Periodic agent-state based Q-learning for POMDPs (NeurIPS 2024 · similarity 0.84)
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.
