Surprise Masking: Private Data Filtering with Model Likelihoods

Reported by arXiv cs.CR (Riya Tyagi, Sahra Ghalebikesabi, Vinnie Monaco, Avi Schwarzschild, Om Thakkar)

Summary by Cyber Security News19 Oct 2026, 06:00 CET

About this summary. This is a short excerpt from the description that arXiv cs.CR published with its article. Cyber Security News did not report or verify the underlying story. Read the original: https://arxiv.org/abs/2610.22988

arXiv:2610.22988v1 Announce Type: new Abstract: Data useful for language model (LM) training can contain private or unwanted information that is at risk of being learned or memorized. We propose Surprise Masking (SM), a simple form of perplexity filtering that uses a publicly available reference model to identify unwanted spans and suppress learning from them while preserving the rest of each example.

Read the full article at arXiv cs.CR →

Source attribution: headline and facts are from arXiv cs.CR (arxiv.org). Summary method: excerpt of the source description. See our source attribution policy.

Related coverage

Misconfigured Supabase apps expose data in over 16,000 databases

AU: Northern Territory Health backtracks on claim only five patients impacted by RDH privacy breach

American Organizations Under Attack: Skyw4lker Device Code Phishing Campaign Exposed

Data breach at Sony-owned internet provider compromises 2.6 million customer records

Pricing Cross-Bank Context in Federated Anti-Money Laundering

Privately Clipping Heavy-Tailed Data