Embryology of a Language Model
Reviewed by Pithpith:USSYQSOHopen to challenge →
read the original abstract
Understanding how language models develop their internal computational structure is a central problem in the science of deep learning. While susceptibilities, drawn from statistical physics, offer a promising analytical tool, their full potential for visualizing network organization remains untapped. In this work, we introduce an embryological approach, applying UMAP to the susceptibility matrix to visualize the model's structural development over training. Our visualizations reveal the emergence of a clear ``body plan,'' charting the formation of known features like the induction circuit and discovering previously unknown structures, such as a ``spacing fin'' dedicated to counting space tokens. This work demonstrates that susceptibility analysis can move beyond validation to uncover novel mechanisms, providing a powerful, holistic lens for studying the developmental principles of complex neural networks.
This paper has not been read by Pith yet.
Forward citations
Cited by 4 Pith papers
-
Mechanistic Anomaly Detection via Functional Attribution
Functional attribution with influence functions detects anomalous mechanisms in neural networks, achieving SOTA backdoor detection (average DER 0.93) on vision benchmarks and improvements on LLMs.
-
Mechanistic Anomaly Detection via Functional Attribution
A grid-free attention-based deep learning model interpolates surface-level PM2.5 concentrations across the US using sparse sensor data and auxiliary geospatial features.
-
Susceptibilities and Patterning: A Primer on Linear Response in Bayesian Learning
Susceptibilities defined via posterior covariances serve as the Jacobian for mapping data distributions to structural coordinates in Bayesian learning, with pseudo-inverse solving for desired structural changes.
-
There Will Be a Scientific Theory of Deep Learning
A mechanics of the learning process is emerging in deep learning theory, characterized by dynamics, coarse statistics, and falsifiable predictions across idealized settings, limits, laws, hyperparameters, and universa...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.