Entities · Models
Pythia
10 articles tagged with this entity.
-
Mechanistic Data Attribution: Tracing the Training Origins of Interpretable LLM Units
-
Function-Vector Heads Are Two Populations: Writers and Cancellers in In-Context Learning
-
Many Circuits, One Mechanism: Input Variation and Evaluation Granularity in Circuit Discovery
-
Trajectory Dynamics in Language Model Hidden States Predict Human Processing Costs Beyond Surprisal
-
BLISS: A Lightweight Bilevel Influence Scoring Method for Data Selection in Language Model Pretraining
-
Cross-Layer Subspace Coupling for LLM Compression: A Unifying Framework and Its Empirical Limits
-
Combating Data Laundering in LLM Training
-
Feature Lottery? A Bifurcation Theory of Concept Emergence
-
Every Component is a Lookup: Token Attribution and Composition from a Single Decomposition
-
LLMs as Noisy Channels: A Shannon Perspective on Model Capacity and Scaling Laws