Benign-trained probes miss tool-action errors under adversarial conditions
Scalar uncertainty beat the probe on AUROC and average precision for Qwen2.5-14B-Instruct under system override.
42
AI model
GLM-4.6V-Flash
- Mentions · 7d1
- Mentions · 30d1
- All mentions1
- First seenNov 8, 2026
- Last mentioned2d ago
Timeline
1 articles · newest firstAppears with
- Qwen2.5-7B-Instructai model · 1
- Qwen2.5-14B-Instructai model · 1
- AgentDojoproduct · 1
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.