Data Curation Matters: Model Collapse and Spurious Shift Performance Prediction from Training on Uncurated Text Embeddings

Training models on uncurated text embeddings (TEs) derived from raw tabular data can lead to a severe failure mode known as $\underline{\text{model collapse}}$, where predictions converge to a single class regardless of input. By comparing models trained with identical hyper-parameter configurations on both raw tabular data and their TE-derived counterparts, we find that collapse is a consistent failure mode in the latter setting. We introduce a set of metrics that capture the extent of model collapse, offering a new perspective on TE quality as a proxy for data curation. Our results reveal that TE alone does not effectively function as a curation layer—and that their quality significantly influences downstream learning. More insidiously, we observe that the presence of model collapse can yield artificially inflated and $\underline{\text{spurious Accuracy-on-the-Line correlation}}$. These findings highlight the need for more nuanced curation and evaluation of embedding-based representations, particularly in out-of-distribution settings.