Interpreting Context Usage in Generative Language Models with Inseq, PECoRe and MIRAGE | Gabriele Sarti
Home
About me
Publications
Blog
Talks
Projects
Activities
CV
Communities
AI2S
AISIG
Interpreting Context Usage in Generative Language Models with Inseq, PECoRe and MIRAGE
Gabriele Sarti
Natural Language Processing
,
Academic
Code
Project
Project
Slides
Date
Jul 16, 2024
Event
CIS LMU Seminar
Location
Ludwig Maximilian University of Munich, Bayern, Germany
Munich, Bavaria, Germany
Natural Language Processing
Interpretability
Sequence-to-sequence
Language Modeling
Feature Attribution
Retrieval-augmented Generation
Related
Interpreting Context Usage in Generative Language Models with Inseq and PECoRe
Post-hoc Interpretability for Generative Language Models: Explaining Context Usage in Transformers
Post-hoc Interpretability for Language Models
Post-hoc Interpretability for NLG & Inseq: an Interpretability Toolkit for Sequence Generation Models
Explaining Neural Language Models from Internal Representations to Model Predictions
Cite
×