Cumulative reward and trap diagnostics on Certified
172,435.94Deceptive RewardDeceptive Leader Mechanism
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Deceptive Leader MechanismDefense strategy=EXP3, Reversion policy=revert, Exploration parameter c=0.2, Honeypot sample constraint=1000-target-sample2026.06 | 172,435.94 | — | 82,764.91 | 95.349 | 1,056.2 | 1,070.2 | |
| Deceptive Leader MechanismDefense strategy=UCB, Reversion policy=revert, Exploration parameter c=0.2, Honeypot sample constraint=1000-target-sample2026.06 | 150,187.61 | — | 55,178.93 | 82.936 | 1,002 | 4,528 | |
| Deceptive Leader MechanismDefense strategy=CP, Reversion policy=revert, Exploration parameter c=0.2, Honeypot sample constraint=1000-target-sample2026.06 | 94,481.64 | — | -527.04 | 52.471 | 1,002 | 1,005 | |
| Deceptive Leader MechanismDefense strategy=UCB, Reversion policy=no revert, Exploration parameter c=0.2, Honeypot sample constraint=1000-target-sample2026.06 | 4,950 | — | -90,058.68 | 2.975 | 1,002 | 4,528 | |
| BaselineDefense strategy=UCB, Reversion policy=no revert, Exploration parameter c=0.2, Honeypot sample constraint=1000-target-sample2026.06 | — | 95,008.68 | — | — | — | — | |
| BaselineDefense strategy=UCB, Reversion policy=revert, Exploration parameter c=0.2, Honeypot sample constraint=1000-target-sample2026.06 | — | 95,008.68 | — | — | — | — | |
| BaselineDefense strategy=CP, Reversion policy=revert, Exploration parameter c=0.2, Honeypot sample constraint=1000-target-sample2026.06 | — | 95,008.68 | — | — | — | — | |
| BaselineDefense strategy=EXP3, Reversion policy=revert, Exploration parameter c=0.2, Honeypot sample constraint=1000-target-sample2026.06 | — | 89,671.03 | — | — | — | — |