Research map / Bandits and online learning / Multi-armed bandits
The Intrinsic Robustness of Stochastic Bandits to Strategic Manipulation
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 |
|---|---|---|---|
| Allow arbitrary adaptive reward manipulation under UCB | UCB · Stochastic bandits | regret: ∑_{i≠i*} [max{3B_i, 81σ²ln T/Δ_i} + (1 + 3Δ_i)] | Mixed |
| Allow arbitrary adaptive reward manipulation under ε-Greedy | ε-Greedy · Stochastic bandits | regret: ∑_{i ≠ i*}[3B_i + O(ln T/Δ_i)] | Mixed |
| Allow arbitrary adaptive reward manipulation under Thompson Sampling | Thompson Sampling · Stochastic bandits | regret: ∑_{i≠i*} max{6B_i, 72σ²ln T/Δ_i} + O(ln T/Δ_i) | Mixed |
| Have strategic arms use LSI | UCB · Stochastic bandits | regret: underline{Δ}∑_{i≠i*} B_i/(2Δ_i) − O(ln T/underline{Δ}) | Hurt |
| Have strategic arms use LSI | ε-Greedy and Thompson Sampling · Stochastic bandits | regret: underline{Δ}∑_{i≠i*} B_i/(2Δ_i) − O(ln T/underline{Δ}) | Hurt |
| Use LSI rather than any other adaptive manipulation strategy | UCB strategic arms · UCB stochastic game | expected number of pulls | Helped |
| Use LSIBR rather than LSI for bounded rewards | Any stochastic bandit algorithm · Bounded rewards | regret: O(underline{Δ}ln T/(1−μ_{i*})²) | Mixed |
| Increase the total manipulation budget | UCB, ε-Greedy, and Thompson Sampling · Simulation, T=10^4 | regret | Hurt |
Closest papers by meaning
- Robust Performance Incentivizing Algorithms for Multi-Armed Bandits with Strategic Agents (AAAI 2025 · similarity 0.91)
- Strategic Multi-Armed Bandit Problems Under Debt-Free Reporting (NeurIPS 2024 · similarity 0.91)
- Nash Social Welfare for Multi Armed Bandits: Trajectory-wise Expected and High Probability Regret (NeurIPS 2026 · similarity 0.90)
- Stochastic Bandits Robust to Adversarial Attacks (ICLR 2025 · similarity 0.89)
- Bandits Meet Mechanism Design to Combat Clickbait in Online Recommendation (ICLR 2024 · similarity 0.89)
- Bandits with Ranking Feedback (NeurIPS 2024 · similarity 0.89)
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.