A technical note on bilinear layers for interpretability
Paper • 2305.03452 • Published • 1
Interpretability
-
-
Interpreting Transformer's Attention Dynamic Memory and Visualizing the Semantic Information Flow of GPT
Paper • 2305.13417 • Published • 1 -
Explainable AI for Pre-Trained Code Models: What Do They Learn? When They Do Not Work?
Paper • 2211.12821 • Published • 2 -
The Linear Representation Hypothesis and the Geometry of Large Language Models
Paper • 2311.03658 • Published • 1 -
Interpreting Pretrained Language Models via Concept Bottlenecks
Paper • 2311.05014 • Published • 1 -
White-Box Transformers via Sparse Rate Reduction
Paper • 2306.01129 • Published • 1 -
ICICLE: Interpretable Class Incremental Continual Learning
Paper • 2303.07811 • Published • 1 -
Differentiable Model Selection for Ensemble Learning
Paper • 2211.00251 • Published • 1 -
CAT-probing: A Metric-based Approach to Interpret How Pre-trained Models for Programming Language Attend Code Structure
Paper • 2210.04633 • Published • 1 -
Forms of Understanding for XAI-Explanations
Paper • 2311.08760 • Published • 2 -
Schema-learning and rebinding as mechanisms of in-context learning and emergence
Paper • 2307.01201 • Published • 2 -
Concept-Centric Transformers: Enhancing Model Interpretability through Object-Centric Concept Learning within a Shared Global Workspace
Paper • 2305.15775 • Published • 1 -
Causal Analysis for Robust Interpretability of Neural Networks
Paper • 2305.08950 • Published • 1 -
Emergence of Segmentation with Minimalistic White-Box Transformers
Paper • 2308.16271 • Published • 17 -
White-Box Transformers via Sparse Rate Reduction: Compression Is All There Is?
Paper • 2311.13110 • Published • 2 -
LLM360: Towards Fully Transparent Open-Source LLMs
Paper • 2312.06550 • Published • 57 -
Patchscope: A Unifying Framework for Inspecting Hidden Representations of Language Models
Paper • 2401.06102 • Published • 21 -
Attention Lens: A Tool for Mechanistically Interpreting the Attention Head Information Retrieval Mechanism
Paper • 2310.16270 • Published • 1