News

2026

Our paper When Does Subspace Direction Matter for LoRA? Regime Analysis of the Magnitude Principle in Few-Shot Adaptation has been accepted to NeurIPS 2026. Congratulations to Nischal!
Welcome Ira to our research group!
Our lab has been approved to join Anthropic’s Claude Team plan for Scientists. Thank you to Anthropic for supporting our research!
New blog post: What Is Negative-Shifted Gradient Descent—When the Useful Estimator Is Not an Endpoint — how a finite-time signed path passes smoothly through the would-be endpoint pole and creates head anti-shrinkage with controlled lower-spectrum exposure.
New blog post: The Spectrum Was Known to Be Biased. Why Did PCR Keep Inverting It? — de-floored PCR and correcting the spectral floor before inversion.
New blog post: Shrinkage Is Not a Universal Law — when expansion, rather than shrinkage, is the right correction in overparameterized prediction.
New preprint: Beyond Negative-Ridge Endpoints: Mixed-Sign Spectral Regularization via Negative-Shifted Gradient Descent.
New preprint: De-floored Principal Component Regression: When Rank Selection Alone Is Insufficient for Prediction.
Our paper An Approximate Bayesian Approach to Covariate-dependent Graphical Modeling was accepted by Bernoulli.
Presented at and chaired a session for the 2026 ICSA Applied Statistics Symposium in Arlington, Virginia.