Shortcut Sensitivity — AgentHarm TS-Guard — Equal interaction weight (197) — F1: leaderboard
Appendix D Table 6 released TS-Guard predictions on AgentHarm-Traj, 731 steps / 197 interactions. Explicit weighting condition: Equal interaction weight (197). No hosted calls made. This is classification of supplied action records; tool calls are not executed. Distinct weighting protocols are not pooled or maximized. Heuristic lookup-rule and marker sensitivities elsewhere are not foundation model scores. Metric: F1. Higher is better. Source: https://arxiv.org/html/2610.22243v1
Metric: F1. Source: arxiv.org. 1 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | TS-Guard 7.6B | 0.72 |
Interactive version: theaggregate.ai/benchmark?slug=shortcut-sensitivity-agentharm-ts-guard-equal-interaction-weight-197-f1 · How It Works · Data refreshed daily, snapshot 2026-10-17.