10 papers
Understanding Contextual Recall in Transformers: How Finetuning Enables In-Context Reasoning over Pretraining Knowledge
Bhavya Vasudeva, Puneesh Deora, Alberto Bietti +2
Transformer-based language models excel at in-context learning (ICL), where they can adapt to new tasks based on contextual examples, without parameter updates. In a specific form…
Why Loss Re-weighting Works If You Stop Early: Training Dynamics of Unconstrained Features
Yize Zhao, Christos Thrampoulidis
The application of loss reweighting in modern deep learning presents a nuanced picture. While it fails to alter the terminal learning phase in overparameterized deep neural network…
How Muon's Spectral Design Benefits Generalization: A Study on Imbalanced Data
Bhavya Vasudeva, Puneesh Deora, Yize Zhao +2
The growing adoption of spectrum-aware matrix-valued optimizers such as Muon and Shampoo in deep learning motivates a systematic study of their generalization properties and, in pa…
Facts in Stats: Impacts of Pretraining Diversity on Language Model Generalization
Tina Behnia, Puneesh Deora, Christos Thrampoulidis
Language models are pretrained on sequences that blend statistical regularities (making text fluent) with factual associations between specific tokens (knowledge of facts). While r…
Geometry of Semantics in Next-Token Prediction: How Optimization Implicitly Organizes Linguistic Representations
Yize Zhao, Christos Thrampoulidis
We investigate how next-token prediction (NTP) optimization leads language models to extract and organize semantic structure from text. Our analysis, based on a tractable mathemati…
In-Context Occam's Razor: How Transformers Prefer Simpler Hypotheses on the Fly
Puneesh Deora, Bhavya Vasudeva, Tina Behnia +1
In-context learning (ICL) enables transformers to adapt to new tasks through contextual examples without parameter updates. While existing research has typically studied ICL in fix…