Research map / Bandits and online learning / Contextual bandits and Thompson sampling
[Re] Bandit Theory and Thompson Sampling-guided Directed Evolution for Sequence Optimization
Luka Žontar
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 |
|---|---|---|---|
| Replace basic DE (random mutation, noisy-evaluation selection) with TS-DE (posterior-guided crossover and mutation) | directed evolution · simulated binary sequences, d=10 | convergence speed of population fitness | Improved |
| Vary mutation rate scheduling of basic DE (ablation) | basic DE · simulated binary sequences | convergence | Hurt |
| Increase population size M from 10 to 90 (population-averaged regret) (ablation) | TS-DE · simulation d=10, T=100, mu=0.8 | population-averaged Bayesian regret | No clear effect |
| Vary directed mutation rate mu in TS-DE (ablation) | TS-DE · simulated fitness curves | fitness over rounds | Improved |
| Run TS-DE with posterior-guided optimization on CRISPR design | TS-DE · CRISPR wet-lab experiments | efficiency fold improvement: 30+ fold | Improved |
Closest papers by meaning
- Bandit Theory and Thompson Sampling-Guided Directed Evolution for Sequence Optimization (NeurIPS 2022 · similarity 0.91)
- Thompson Sampling via Fine-Tuning of LLMs (ICLR 2026 · similarity 0.85)
- Tree Search-Based Evolutionary Bandits for Protein Sequence Optimization (AAAI 2024 · similarity 0.84)
- Improved Off-policy Reinforcement Learning in Biological Sequence Design (ICML 2025 · similarity 0.84)
- Steering Generative Models with Experimental Data for Protein Fitness Optimization (NeurIPS 2025 · similarity 0.83)
- Teacher-Aware Evolution of Heuristic Programs from Learned Optimization Policies (NeurIPS 2026 · similarity 0.82)
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.
