Rewarded soups: towards Pareto-optimal alignment by interpolating weights fine-tuned on diverse rewards
Paper
•
2306.04488
•
Published
•
2
Corentin Dancette
cdancette
·
AI & ML interests
Multimodal learning, biases, OOD generalization
Organizations
authored
4
papers
about 1 year ago
eP-ALM: Efficient Perceptual Augmentation of Language Models
Paper
•
2303.11403
•
Published
•
3
Beyond Task Performance: Evaluating and Reducing the Flaws of Large Multimodal Models with In-Context Learning
Paper
•
2310.00647
•
Published
Efficient Medical Question Answering with Knowledge-Augmented Question Generation
Paper
•
2405.14654
•
Published
•
1
authored
a
paper
about 2 years ago