finding
active
finding:verbalization-monitor-asymptotes-at-auroc-0-75-even-at-100-rollouts-on-fortress-vs-wildchatVerbalization monitor asymptotes at AUROC 0.75 even at 100 rollouts on Fortress vs WildChat
Baseline ceiling.
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Logit monitor reaches AUROC 0.79 from a single rollout separating Fortress from WildChat on Kimi K2.5finding0.844Key result.
- Qualitative picture in Figure 1.
- On Qwen 3 32B with Fortress, logit monitor at one rollout outperforms verbalization monitor at 100 rolloutsfinding0.799Additional model replication.
- Prediction task result.
- Prompt-level effect is much larger.
- Additional dataset replication.
- Rollout-level verbalization barely matters.
- Robustness check.