research
∙
09/05/2023
Provably safe systems: the only path to controllable AGI
We describe a path to humanity safely thriving with powerful Artificial ...
share
research
∙
06/30/2023
The Clock and the Pizza: Two Stories in Mechanistic Explanation of Neural Networks
Do neural networks, trained on well-understood algorithmic tasks, reliab...
research
∙
05/31/2023
Discovering New Interpretable Conservation Laws as Sparse Invariants
Discovering conservation laws for a given dynamical system is important ...
research
∙
05/04/2023
Seeing is Believing: Brain-Inspired Modular Training for Mechanistic Interpretability
We introduce Brain-Inspired Modular Training (BIMT), a method for making...
research
∙
04/05/2023
GenPhys: From Physical Processes to Generative Models
Since diffusion models (DM) and the more recent Poisson flow generative ...
research
∙
03/23/2023
The Quantization Model of Neural Scaling
We propose the Quantization Model of neural scaling laws, explaining bot...
research
∙
02/08/2023
PFGM++: Unlocking the Potential of Physics-Inspired Generative Models
We introduce a new family of physics-inspired generative models termed P...
research
∙
10/24/2022
Precision Machine Learning
We explore unique considerations involved in fitting ML models to data w...
research
∙
10/03/2022
Omnigrok: Grokking Beyond Algorithmic Data
Grokking, the unusual phenomenon for algorithmic datasets where generali...
research
∙
09/22/2022
Poisson Flow Generative Models
We propose a new "Poisson flow" generative model (PFGM) that maps a unif...
research
∙
05/20/2022
Towards Understanding Grokking: An Effective Theory of Representation Learning
We aim to understand grokking, a phenomenon where models generalize long...
research
∙
04/05/2022
Pareto-optimal clustering with the primal deterministic information bottleneck
At the heart of both lossy compression and clustering is a trade-off bet...
research
∙
03/23/2022
AI Poincaré 2.0: Machine Learning Conservation Laws from Differential Equations
We present a machine learning algorithm that discovers conservation laws...
research
∙
02/25/2022
Biological error correction codes generate fault-tolerant neural networks
It has been an open question in deep learning if fault-tolerant computat...
research
∙
09/28/2021
Physics-Augmented Learning: A New Paradigm Beyond Physics-Informed Learning
Integrating physical inductive biases into machine learning can improve ...
research
∙
09/20/2021
Machine-learning hidden symmetries
We present an automated method for finding hidden symmetries, defined as...
research
∙
08/31/2021
Machine-Learning media bias
We present an automated method for measuring media bias. Inferring which...
research
∙
05/31/2021
Machine-Learning Non-Conservative Dynamics for New-Physics Detection
Energy conservation is a basic physics principle, the breakdown of which...
research
∙
11/09/2020
AI Poincaré: Machine Learning Conservation Laws from Trajectories
We present AI Poincaré, a machine learning algorithm for auto-discoverin...
research
∙
06/18/2020
AI Feynman 2.0: Pareto-optimal symbolic regression exploiting graph modularity
We present an improved method for symbolic regression that seeks to fit ...
research
∙
05/19/2020
Symbolic Pregression: Discovering Physical Laws from Raw Distorted Video
We present a method for unsupervised learning of equations of motion for...
research
∙
08/23/2019
Pareto-optimal data compression for binary classification tasks
The goal of lossy data compression is to reduce the storage cost of a da...
research
∙
07/17/2019
Learnability for the Information Bottleneck
The Information Bottleneck (IB) method (tishby2000information) provides ...
research
∙
05/27/2019
AI Feynman: a Physics-Inspired Method for Symbolic Regression
A core challenge for both physics and artificial intellicence (AI) is sy...
research
∙
04/30/2019
The role of artificial intelligence in achieving the Sustainable Development Goals
The emergence of artificial intelligence (AI) and its progressively wide...
research
∙
02/09/2019
Optimal Latent Representations: Distilling Mutual Information into Principal Pairs
Principal component analysis (PCA) is generalized from one to two random...
research
∙
10/24/2018
Toward an AI Physicist for Unsupervised Learning
We investigate opportunities and challenges for improving unsupervised m...
research
∙
07/26/2018
Meta-learning autoencoders for few-shot prediction
Compared to humans, machine learning models generally require significan...
research
∙
06/08/2017
Gated Orthogonal Recurrent Units: On Learning to Forget
We present a novel recurrent neural network (RNN) based model that combi...
research
∙
05/16/2017
The power of deeper networks for expressing natural functions
It is well-known that neural networks are universal approximators, but t...
research
∙
12/15/2016
Tunable Efficient Unitary Neural Networks (EUNN) and their application to RNNs
Using unitary (instead of general) matrices in artificial neural network...
research
∙
08/29/2016
Why does deep and cheap learning work so well?
We show how the success of deep learning could depend not only on mathem...
research
∙
06/21/2016
Criticality in Formal Languages and Statistical Physics
We show that the mutual information between two symbols, as a function o...
research
∙
02/10/2016
Research Priorities for Robust and Beneficial Artificial Intelligence
Success in the quest for artificial intelligence has the potential to br...
research
∙
09/02/2014
Friendly Artificial Intelligence: the Physics Challenge
Relentless progress in artificial intelligence (AI) is increasingly rais...
research
∙
07/05/1999