Newest papers

The Effectiveness and Durability of Pretraining Data Filtering in Protein Language Models

Aaron Maiwald, Roderick McNeill, 이재영, Matthew A. Clarke, Lennart J. Justen, Alejandro Acelas, Ivo Andrews, Giles Harper-Donnelly, Adam Moore, Sophie Rose, Oliver M. Crook, Felix Jackson · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.26651 · cited by 0 works (OpenAlex) · open-access PDF

Hindi

प्री-ट्रेनिंग डेटा फ़िल्टरिंग ओपन-वेग भाषा मॉडल के लिए सबसे आम सुरक्षा उपायों में से एक है। यहाँ हम प्रोटीन भाषा मॉडल के ESM-2 परिवार के साथ इसकी प्रभावशीलता, टिकाऊता और स्केलेबलता का अध्ययन करते हैं। हम पाते हैं कि प्री-ट्रेनिंग डेटा से वायरल प्रोटीन फ़िल्टरिंग मायने में मास्क भाषा मॉडलिंग, म्यूटेशन प्रभाव पूर्वानुमान और वायरल प्रोटीन पर लंबी दूरी के संपर्क पूर्वानुमान को कम करता है, जबकि मुख्य रूप से गैर-वायरल प्रोटीन पर प्रदर्शन को बनाए रखता है। इन प्रभावों को मानव संक्रमित वायरसों से प्रोटीनों के बीच केंद्रित किया जाता है, और हम 8M से 3B पैरामीटरों

Machine translation (Meta M2M100); the English original follows.

Abstract

Pretraining filtering is one of the most common safeguards for open-weight language models. Here we study its effectiveness, durability and scalability using the ESM-2 family of protein language models. We find that filtering viral proteins from the pretraining data meaningfully degrades masked language modelling, mutation effect prediction and long-range contact prediction on viral proteins, while largely preserving performance on non-viral proteins. These effects are concentrated among proteins from human-infecting viruses, and grow as we scale models from 8M to 3B parameters. As models increase in size, we find that stringent filtering becomes more effective. However, fine-tuning on sequences from the removed viral families restores the lost performance. Although the cost to recover performance increases with model scale, even for the largest models it costs about 80 USD of H100 time. This work establishes pretraining filtering as moderately effective at selectively degrading performance on viral proteins, but not durable against fine-tuning with modest computational resources and access to relevant data.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Scalability, Computer science, Language model, Durability, Artificial intelligence, Scale (ratio)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange