Research map / Bandits and online learning / Online convex optimization and games
CLIP-OGD: An Experimental Design for Adaptive Neyman Allocation in Sequential Experiments
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 |
|---|---|---|---|
| use CLIP-OGD instead of Bernoulli design for confidence intervals (Chebyshev width) | adaptive Horvitz-Thompson estimator · Groh-McKenzie simulation, T=14,445 | Chebyshev CI width: 0.0541 -> 0.0507 | Helped |
| use CLIP-OGD instead of Bernoulli design for normal-based intervals (width) | adaptive Horvitz-Thompson estimator · Groh-McKenzie simulation, T=14,445 | Normal CI width: 0.0237 -> 0.0222 | Helped |
| use CLIP-OGD instead of Bernoulli (normal-based coverage) | adaptive Horvitz-Thompson estimator · Groh-McKenzie simulation, T=14,445 | Normal CI coverage: 95.21% -> 95.22% | No clear effect |
| use Chebyshev intervals instead of normal-based intervals | CLIP-OGD · Groh-McKenzie simulation, T=14,445 | CI coverage: 99.99% vs 95.22% | Mixed |
| use Explore-then-Commit instead of CLIP-OGD in large samples | adaptive Neyman allocation · Groh-McKenzie simulation, original dataset | normalized variance | Hurt |
| flip potential outcomes of first 100 units, applied to CLIP-OGD (ablation) | CLIP-OGD · Groh-McKenzie simulation, flipped outcomes | normalized variance | No clear effect |
| flip potential outcomes of first 100 units, applied to ETC (ablation) | Explore-then-Commit · Groh-McKenzie simulation, flipped outcomes | normalized variance | Hurt |
| use smaller step size c in eta=c/sqrt(T) for CLIP-OGD (ablation) | CLIP-OGD · Groh-McKenzie simulation, step-size sweep c in {0.25,0.5,1.0,2.0,4.0} | normalized variance convergence rate | Helped |
| use DBCD and EXP3 instead of CLIP-OGD | adaptive designs · Groh-McKenzie simulation, unnormalized variance | variance | Hurt |
Closest papers by meaning
- Stronger Neyman Regret Guarantees for Adaptive Experimental Design (ICML 2025 · similarity 0.94)
- Efficient Adaptive Experimentation with Noncompliance (NeurIPS 2025 · similarity 0.88)
- Optimistic Algorithms for Adaptive Estimation of the Average Treatment Effect (ICML 2025 · similarity 0.87)
- Active Adaptive Experimental Design for Treatment Effect Estimation with Covariate Choice (ICML 2024 · similarity 0.86)
- Near-Optimal Experiment Design in Linear non-Gaussian Cyclic Models (NeurIPS 2025 · similarity 0.86)
- Simulation-Based Inference for Adaptive Experiments (NeurIPS 2025 · similarity 0.86)
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.