DELTA-CROSSCODER: ROBUST CROSSCODER IN NARROW FINE-TUNING REGIMES

Aly M. Kassem, Thomas Jiralerspong, Negar Rostamzadeh, Golnoosh Farnadi

Keywords: model diffing, crosscoders, narrow fine-tuning, interpretability, sparse autoencoders

Abstract: Model diffing methods aim to identify how fine-tuning changes a model's internal representations. Crosscoders approach this by learning shared dictionaries of interpretable latent directions between base and fine-tuned models. However, existing formulations struggle with narrow fine-tuning, where behavioral changes are localized and asymmetric. We introduce Delta-Crosscoder, which combines Dual-K BatchTopK sparsity with a delta-based loss prioritizing directions that change between models, plus an implicit contrastive signal from paired activations on matched inputs. Evaluated across synthetic false facts, emergent misalignment, subliminal learning, and taboo word games (Gemma, LLaMA, Qwen; 1B–7B parameters), Delta-Crosscoder reliably isolates latent directions causally responsible for fine-tuned behaviors and enables effective mitigation, substantially outperforming baselines. Our results demonstrate that narrow fine-tuning induces distinctive, recoverable latent shifts and that crosscoder methods remain powerful tools for model diffing.

Submission Number: 258