conftrace_

linear feature

5 papers

Explore in graph

Co-occurring keywords · papers mentioning both

large language model (1) selective question answering (1) linear-predictive feature (1) truth direction (1) trace of hessian (1) bias (1) observable propagation (1) label noise sgd (1) rank one feature matrix (1) feature vector (1)

Papers

Probing the Geometry of Truth: Consistency and Generalization of Truth Directions in LLMs Across Logical Transformations and Question Answering Tasks ACL 2025 Observable Propagation: Uncovering Feature Vectors in Transformers ICML 2024 Simplicity Bias via Global Convergence of Sharpness Minimization ICML 2024 Evading the Simplicity Bias: Training a Diverse Set of Models Discovers Solutions With Superior OOD Generalization CVPR 2022 Sample-Optimal Parametric Q-Learning Using Linearly Additive Features ICML 2019