4 papers · 1 filter
Latent Chain-of-Thought Improves Structured-Data Transformers
Carson Dudley, Samet Oymak
Chain-of-thought and more broadly test-time compute are known to augment the expressive capabilities of language models and have led to major innovations in reasoning. Motivated by…
In-Context Learning Under Regime Change
Carson Dudley, Yutong Bi, Xiaofeng Liu +1
Non-stationary sequences arise naturally in control, forecasting, and decision-making. The data-generating process shifts at unknown times, and models must detect the change, disca…
Simulation as Supervision: Mechanistic Pretraining for Scientific Discovery
Carson Dudley, Reiden Magdaleno, Christopher Harding +1
Scientific modeling faces a tradeoff between the interpretability of mechanistic theory and the predictive power of machine learning. While existing hybrid approaches have made pro…
Learning From Simulators: A Theory of Simulation-Grounded Learning
Carson Dudley, Marisa Eisenberg
Simulation-Grounded Neural Networks (SGNNs) are predictive models trained entirely on synthetic data from mechanistic simulations. They have achieved state-of-the-art performance i…