Patrick Lewis
multilingual
efficiency
multilinguality
safety
adaptation
large language model
pre-trained language models
llm
sequence labeling
instruction tuning
neural retrieval
toxicity
efficient nlp
information retrieval
cross-lingual
6
presentations
4
number of views
SHORT BIO
Bio: Patrick Lewis is a Research Scientist at Meta AI in London. Patrick recently completed his PhD at UCL and FAIR, under the supervision of Sebastian Riedel and Pontus Stenetorp. Patrick has worked extensively in Question Answering, and more broadly in Knowledge-intensive NLP – NLP tasks that require a substantial amount of world knowledge to do well on, and which an average human would require access to a search engine, library or external knowledge source to do well. Patrick is interested in how to build and evaluate models that precisely access, apply and attribute knowledge, in terms of accuracy, updateability, controllability and interpretability.
Presentations
From One to Many: Expanding the Scope of Toxicity Mitigation in Language Models
Beyza Ermis and 3 other authors
On the Challenges of Using Black-Box APIs for Toxicity Evaluation in Research | VIDEO
Luiza Amador Pozzobon and 3 other authors
Mini-Model Adaptation: Efficiently Extending Pretrained Models to New Languages via Aligned Shallow Training
Kelly Marchisio and 3 other authors
Mini-Model Adaptation: Efficiently Extending Pretrained Models to New Languages via Aligned Shallow Training
Kelly Marchisio and 3 other authors
Task-aware Retrieval with Instructions
Akari Asai and 7 other authors
Boosted Dense Retriever
Patrick Lewis and 5 other authors