Machine Learning Posts

Stanford AI Lab Papers and Talks at ICML 2023

All the great work from the Stanford AI Lab accepted at ICML 2023, all in one place.

Self-Improving Robots: Embracing Autonomy in Robot Learning

Learning methods for robotics have a data problem. Autonomous robotic systems can deliver orders of magnitude more real-world embodied data than we have.

From Discrimination in Machine Learning to Discrimination in Law, Part 1: Disparate Treatment

Fereshte Khani, Percy Liang

Learning to Imitate

Divyansh Garg

SOTA methods for Imitation Learning are heavily data-inefficient or rely on unstable Adversarial training. We present a novel framework that introduces simple, stable, and data-efficient learning from few experts that scales well to complex environments.

Can Longer Sequences Help Take the Next Leap in AI?

Chris Ré, Tri Dao, Dan Fu, Karan Goel

Why we think modeling longer sequences is exciting, and highlights of our recent work advancing this task.

LinkBERT: Improving Language Model Training with Document Link

Michihiro Yasunaga, Jure Leskovec, Percy Liang

LinkBERT is a new language model pretrained to capture document link knowledge such as hyperlinks of the web. It greatly helps knowledge-intensive applications such as question answering.

Understanding Deep Learning Algorithms that Leverage Unlabeled Data, Part 2: Contrastive Learning

Jeff Z. HaoChen, Colin Wei, and Tengyu Ma

Theoretical analysis of contrastive learning algorithms for leveraging unlabeled data.

Discovering the systematic errors made by machine learning models

Sabri Eyuboglu, Maya Varma, Khaled Saab, Jared Dunnmon, James Zou and Chris Ré

Machine learning models that achieve high overall accuracy often make systematic errors on coherent slices of validation data. Can we develop methods to automatically identify these systematic errors?

Understanding Deep Learning Algorithms that Leverage Unlabeled Data, Part 1: Self-training

Colin Wei, Jeff Z. HaoChen, and Tengyu Ma

Theoretical analysis of self-training algorithms for leveraging unlabeled data.

How to Improve User Experience (and Behavior): Three Papers from Stanford's Alexa Prize Team

Amelia Hardy, Haojun Li, and Abigail See

Strategies for understanding user dissatisfaction, handling offensiveness, and increasing user initiative, to improve conversational experience.