Research map / Probabilistic and scientific modeling / Causal discovery
Structural Causal Bandits: Where to Intervene?
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| restrict arms to POMISs instead of intervening on all variables at once | Thompson sampling · Task 3 | cumulative regret: 2784.8 -> 91.4 | Helped |
| restrict arms to POMISs instead of all-at-once | Thompson sampling · Task 2 (IV) | cumulative regret: 272.1 -> 16.1 | Helped |
| restrict arms to POMISs instead of brute-force over all subsets | Thompson sampling · Task 3 | cumulative regret: 1469.0 -> 91.4 | Helped |
| restrict arms to POMISs instead of MISs (ablation) | Thompson sampling · Task 3 | cumulative regret: 472.4 -> 91.4 | Helped |
| restrict arms to POMISs instead of MISs (ablation) | Thompson sampling · Task 2 (IV) | cumulative regret: 21.4 -> 16.1 | Helped |
| restrict arms to POMISs instead of all-at-once in Markovian model | kl-UCB · Task 1 | cumulative regret: 12 -> 3.0 | Helped |
| use POMIS arm set instead of all-at-once | kl-UCB · Task 1 | rounds to 95% optimal arm probability: 66 -> 20 | Helped |
| use MIS instead of brute-force arm set at early stage (ablation) | bandit solver · Task 1 | optimal arm probability | Hurt |
Closest papers by meaning
- Partial Structure Discovery is Sufficient for No-regret Learning in Causal Bandits (NeurIPS 2024 · similarity 0.91)
- Causal Bandits: The Pareto Optimal Frontier of Adaptivity, a Reduction to Linear Bandits, and Limitations around Unknown Marginals (ICML 2024 · similarity 0.91)
- Certified Policy Optimisation for Nested Causal Bandits via PAC-Bayes Risk (NeurIPS 2026 · similarity 0.90)
- Prediction-Intervention Games and Invariant Sets (NeurIPS 2026 · similarity 0.88)
- Adversarial Causal Bayesian Optimization (ICLR 2024 · similarity 0.88)
- Linear Causal Bandits: Unknown Graph and Soft Interventions (NeurIPS 2024 · similarity 0.88)
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.