{"ID":23475193,"CreatedAt":"2026-09-18T01:09:05.407443952Z","UpdatedAt":"2026-09-20T18:11:56.143995915Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2609.20027","arxiv_id":"2609.20027","title":"Can Data Attribution Filter Out Subliminal Learning? Not Reliably","abstract":"Subliminal learning allows language models to transmit behavioral traits through training data with no obvious semantic relationship to those traits, undermining content-based data filtering as a safety intervention. Training data attribution offers an alternative: it identifies the training examples responsible for a given model behavior, independent of their semantic content, and so may apply in exactly the cases where semantic inspection fails. We evaluate three gradient-based attribution methods (GradCos, a contrastive GradCos variant, and EK-FAC) across three models, comparing them against divergence tokens, a strong baseline previously shown to localize subliminal learning (albeit one that requires access to counterfactual teacher models). Filtering at the token level, EK-FAC mitigates a significant part of the effect, the other methods provide little benefit, and all mostly fall short of divergence tokens. Filtering entire samples is less effective for every method, though EK-FAC often gives a stronger signal than divergence tokens in this setting. Success is inconsistent across methods and settings: variants that work well for some model-preference combinations fail for others, and we do not identify a consistent explanation for these differences. Our results suggest that gradient-based attribution can identify data responsible for subliminal learning in some settings, but that some approximations are more reliable than others.","short_abstract":"Subliminal learning allows language models to transmit behavioral traits through training data with no obvious semantic relationship to those traits, undermining content-based data filtering as a safety intervention. Training data attribution offers an alternative: it identifies the training examples responsible for a...","url_abs":"https://arxiv.org/abs/2609.20027","url_pdf":"https://arxiv.org/pdf/2609.20027v1","authors":"[\"Moritz Weckbecker\",\"Sweta Jena\",\"Jonas Müller\",\"Ponnurangam Kumaraguru\",\"Sebastian Lapuschkin\",\"Wojciech Samek\",\"Louis Jaburi\",\"Gonçalo Paulo\"]","published":"2026-09-17T10:34:18Z","proceeding":"cs.AI","tasks":"[\"cs.AI\"]","methods":"[\"Language Model\"]","has_code":false}