Reproduce “Quantifying Memorization and Privacy Risks in Genomic Language Models” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Alexander Nemecek, Wenbiao Li, Xiaoqian Jiang, Jaideep Vaidya, Erman Ayday
The authors audit memorization in genomic language models fine-tuned on small datasets by inserting random DNA canaries and measuring familiarity, extraction, and training-set membership. They combine these measures in a worst-case score. Evo with LoRA recovered every planted canary at one insertion on three real-data datasets, but these controlled probes do not establish leakage rates for genomes or larger cohorts.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.