Metraipolis › archive › dc_e22c4cf4cea4
Scaling Laws for Safety Probes — arXiv:2610.30678
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.30678
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-25T21:34:32.804Z
- SHA-256 of the archived text
83ce43beaa72329f15154bfb642f1c2a26cac067dcceba0a48bcefae3297a340- Stable URL of this page
https://metraipolis.com/documento/dc_e22c4cf4cea4
How to cite
Scaling Laws for Safety Probes — arXiv:2610.30678 — arXiv (en), https://arxiv.org/abs/2610.30678 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e22c4cf4cea4 (Metraipolis archive dc_e22c4cf4cea4, sha256 83ce43beaa72329f15154bfb642f1c2a26cac067dcceba0a48bcefae3297a340)
BibTeX
@misc{dce22c4cf4cea4,
title = {Scaling Laws for Safety Probes — arXiv:2610.30678},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.30678},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e22c4cf4cea4; sha256 83ce43beaa72329f15154bfb642f1c2a26cac067dcceba0a48bcefae3297a340},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.30678. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Safety probes trained on a model's activations are a standard tool for monitoring language models in real time. Each time a new model is released or a new harm area needs to be monitored, the probe must be retrained, which requires new labels and a choice of probe architecture. Yet neither the model nor the labels tell us in advance how well the resulting probe will perform. Finding out requires training and comparing many probe configurations. In this paper, we ask whether probe performance can instead be predicted in advance from model size and the number of labels. Existing safety datasets are insufficient for this, as they are either limited in scale or already saturated. We therefore construct six new safety datasets of sufficient size and difficulty. On these datasets, we train two probe architectures on 20 base models from five model families, with 32 to $10^{6}$ labels for each model. We measure performance using probe error, defined as 1 - AUROC, where lower values are better. We find that probe error follows two simple scaling laws. First, the lowest probe error a model can reach falls as a power law in its size. Second, probe error decreases towards its floor as the number of examples increases; the slope increases as the target model size increases. Both laws hold across tasks, model families and probe architectures, and only their fitted constants differ. When fitted on the smallest models of a family, the laws predict the error of the next larger model to within 0.02 in 72% of cases. They also identify whether more labels or a larger model is the cheaper way to reach a target error in 73% of cases. Finally, they choose the probe architecture with an extra error of only 0.006 on average when they pick the worse one. These laws allow safety researchers to determine how to create the best probe for any given model and label budget.
Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Cryptography and Security (cs.CR); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.30678 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-25T21:25:18.101Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).