Getting Modular with Language Models: Building, Reusing a Library of Experts for Task Generalization

Microsoft Research Forum | Episode 2 | March 5, 2024

Alessandro Sordoni shared recent efforts on building and re-using large collections of expert language models to improve zero-shot and few-shot generalization to unseen tasks.

See more at https://aka.ms/ResearchForum-Mar2024 (opens in new tab)

Séries:: Microsoft Research Forum
Date:: March 5, 2024

- Alessandro Sordoni
  
  Principal Researcher
Domaine de recherche
- Artificial intelligence
Laboratoire de recherche
- Microsoft Research Lab – Montréal
Événement
- Microsoft Research Forum | Episode 2
Publication
- LoRA: Low-Rank Adaptation of Large Language Models
Télécharger
- LoRA
- Phi-2

Taille: Microsoft Research Forum

Using LLMs for safe low-level programming | Microsoft Research Forum
February 25, 2025
Speakers:

Aseem Rastogi,

Pantazis Deligiannis
AutoGen v0.4: Reimagining the foundation of agentic AI for scale and more | Microsoft Research Forum
February 25, 2025
Speakers:

Gagan Bansal
Belief state transformers | Microsoft Research Forum
February 25, 2025
Speakers:

John Langford
Magma: A foundation model for multimodal AI Agents | Microsoft Research Forum
February 25, 2025
Speakers:

Jianwei Yang
Chimera: Accurate synthesis prediction by ensembling models with... | Microsoft Research Forum
February 25, 2025
Speakers:

Marwin Segler
AI for Precision Health: Learning the language of nature and patients | Microsoft Research Forum
February 25, 2025
Speakers:

Hoifung Poon,

Ava Amini,

Lili Qiu

, et. al.
Keynote: Multimodal Generative AI for Precision Health | Microsoft Research Forum
February 25, 2025
Speakers:

Hoifung Poon
Fostering appropriate reliance on AI
September 3, 2024
Speakers:

Mihaela Vorvoreanu
A generative model of biology for in-silico experimentation and discovery
September 3, 2024
Speakers:

Kevin Yang
Project Aurora: The first large-scale foundation model of the atmosphere
September 3, 2024
Speakers:

Megan Stanley
Direct Nash Optimization: Teaching language models to self-improve with general preferences
September 3, 2024
Speakers:

Corby Rosset
Analog optical computing for sustainable AI and beyond
September 3, 2024
Speakers:

Jiaqi Chu
Research Forum 4 | Panel: Beyond Language: The future of multimodal models in healthcare, gaming, and AI
September 3, 2024
Speakers:

John Langford,

Katja Hofmann,

Jianwei Yang

, et. al.
Research Forum 4 | Keynote: Phi-3-Vision: A highly capable and "small" language vision model
September 3, 2024
Speakers:

Jianfeng Gao
AutoGen Update: Complex Tasks and Agents
June 4, 2024
Speakers:

Adam Fourney
MatterGen: A Generative Model for Materials Design
June 4, 2024
Speakers:

Tian Xie
Driving Industry Evolution: Exploring the Impact of Generative AI on Sector Transformation
June 4, 2024
Speakers:

Jiang Bian
Insights into the Challenges and Opportunities of Large Multi-Modal Models for Blind and Low Vision Users: CLIP
June 4, 2024
Speakers:

Daniela Massiceti
Research Forum 3 | Panel: Generative AI for Global Impact: Challenges and Opportunities
June 4, 2024
Speakers:

Jacki O'Neill,

Tanuja Ganu,

Sunayana Sitaram

, et. al.
Research Forum 3 | Keynote: Building Globally Equitable AI
June 4, 2024
Speakers:

Jacki O'Neill
Generative AI and Plural Governance: Mitigating Challenges and Surfacing Opportunities
March 5, 2024
Speakers:

Madeleine Daepp
GigaPath: Foundation Model for Digital Pathology
March 5, 2024
Speakers:

Naoto Usuyama
Getting Modular with Language Models: Building, Reusing a Library of Experts for Task Generalization
March 5, 2024
Speakers:

Alessandro Sordoni
The Metacognitive Demands and Opportunities of Generative AI
March 5, 2024
Speakers:

Lev Tankelevitch
What's new in AutoGen?
March 5, 2024
Speakers:

Chi Wang