Scaling Inherently Interpretable Language Models
Paper • 2608.07594 • Published • 20
AI & ML interests
We build interpretable models and AI systems that can reliably explain their reasoning, and are easy to audit, steer, and understand.
Recent Activity
View all activity
Add chat_template.jinja; drop inline chat_template
1
#3 opened 18 days ago
by 
MuawizC
Add chat_template.jinja; drop inline chat_template
1
#3 opened 18 days ago
by MuawizC
Add links to paper, code, and project page
❤️ 1
1
#7 opened 3 months ago
by 
nielsr
luulinh90s
submitted a
paper to Daily Papers 3 months ago
AyaGL
updated a
model 4 months ago
AyaGL
published a
model 4 months ago
AyaGL
updated 2
models 5 months ago
AyaGL
published a
model 5 months ago
andreasmadsen
authored 6
papers over 2 years ago
Interpretability Needs a New Paradigm
Paper • 2405.05386 • Published • 5
Evaluating the Faithfulness of Importance Measures in NLP by Recursively Masking Allegedly Important Tokens and Retraining
Paper • 2110.08412 • Published • 1
Faithfulness Measurable Masked Language Models
Paper • 2310.07819 • Published
Measuring Arithmetic Extrapolation Performance
Paper • 1910.01888 • Published
Neural Arithmetic Units
Paper • 2001.05016 • Published
Post-hoc Interpretability for Neural NLP: A Survey
Paper • 2108.04840 • Published
andreasmadsen
authored a
paper almost 3 years ago