Search papers, labs, and topics across Lattice.
This paper introduces WARP, a novel framework that analyzes the weight space of fine-tuned models to recover the training data portfolios used in their creation. By interpolating between base and fine-tuned models through model merging, WARP generates pseudo-checkpoints that reveal the geometric characteristics of the training data. In experiments with BERT and GPT-2, WARP achieves an impressive average mean absolute error (MAE) of 0.046 and 0.104 in recovering domain mixtures, significantly outperforming traditional membership inference methods.
WARP reveals the hidden training data portfolios of foundation models with remarkable accuracy, challenging the opacity of model training processes.
Foundation models are routinely released to the public, yet the data recipes used to train them -- such as domain mixture weights that determine how different sources are sampled -- are rarely disclosed. This creates an access asymmetry: researchers study the resulting models but lack visibility into the training distribution that produces them. Prior works for inferring training data, such as membership inference, detect at the level of individual samples and thus cannot characterize the global composition of the training corpus. We introduce WARP, a framework that recovers a fine-tuned model's training mixtures directly from its released weights. WARP interpolates between the base and fine-tuned models using model merging, generating pseudo-checkpoints that approximate the missing training trajectory and expose a geometric footprint of the training data in the weight space. From these simulated footprints, WARP extracts geometric features and maps them to domain proportions using either a parameter-free softmax readout or an MLP projector trained on synthetic mixtures. In controlled experiments with BERT and GPT-2, WARP recovers domain mixtures with an average MAE as low as 0.046 and 0.104 respectively, outperforming membership inference and a variant with access to the true training trajectory.
