Search Results for author: Max Tegmark

Found 52 papers, 23 papers with code

GenEFT: Understanding Statics and Dynamics of Model Generalization via Effective Theory

no code implementations8 Feb 2024 David D. Baek, Ziming Liu, Max Tegmark

We present GenEFT: an effective theory framework for shedding light on the statics and dynamics of neural network generalization, and illustrate it with graph learning examples.

Graph Learning Representation Learning

A Resource Model For Neural Scaling Law

no code implementations7 Feb 2024 Jinyeop Song, Ziming Liu, Max Tegmark, Jeff Gore

A task is usually composite hence can be decomposed into many subtasks, which compete for resources (measured by the number of neurons allocated to subtasks).

Opening the AI black box: program synthesis via mechanistic interpretability

1 code implementation7 Feb 2024 Eric J. Michaud, Isaac Liao, Vedang Lad, Ziming Liu, Anish Mudide, Chloe Loughridge, Zifan Carl Guo, Tara Rezaei Kheirkhah, Mateja Vukelić, Max Tegmark

We present MIPS, a novel method for program synthesis based on automated mechanistic interpretability of neural networks trained to perform the desired task, auto-distilling the learned algorithm into Python code.

Program Synthesis Symbolic Regression

Generating Interpretable Networks using Hypernetworks

no code implementations5 Dec 2023 Isaac Liao, Ziming Liu, Max Tegmark

The hypernetwork is carefully designed such that it can control network complexity, leading to a diverse family of interpretable algorithms ranked by their complexity.

Systematic Generalization

Growing Brains: Co-emergence of Anatomical and Functional Modularity in Recurrent Neural Networks

no code implementations11 Oct 2023 Ziming Liu, Mikail Khona, Ila R. Fiete, Max Tegmark

Recurrent neural networks (RNNs) trained on compositional tasks can exhibit functional modularity, in which neurons can be clustered by activity similarity and participation in shared computational subtasks.


The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets

1 code implementation10 Oct 2023 Samuel Marks, Max Tegmark

In this work, we curate high-quality datasets of true/false statements and use them to study in detail the structure of LLM representations of truth, drawing on three lines of evidence: 1.

Language Modelling Large Language Model

Grokking as Compression: A Nonlinear Complexity Perspective

no code implementations9 Oct 2023 Ziming Liu, Ziqian Zhong, Max Tegmark

To do so, we define linear mapping number (LMN) to measure network complexity, which is a generalized version of linear region number for ReLU networks.

Attribute Memorization +2

Divide-and-Conquer Dynamics in AI-Driven Disempowerment

no code implementations9 Oct 2023 Peter S. Park, Max Tegmark

Myopic members prioritize their future well-being less than their present well-being, and are thus disinclined to solidarily support current victims today at personal cost, even if this is necessary to counter the shared threat of AI-driven disempowerment.

Language Models Represent Space and Time

1 code implementation3 Oct 2023 Wes Gurnee, Max Tegmark

The capabilities of large language models (LLMs) have sparked debate over whether such systems just learn an enormous collection of superficial statistics or a coherent model of the data generation process -- a world model.

A Neural Scaling Law from Lottery Ticket Ensembling

no code implementations3 Oct 2023 Ziming Liu, Max Tegmark

Neural scaling laws (NSL) refer to the phenomenon where model performance improves with scale.


Provably safe systems: the only path to controllable AGI

no code implementations5 Sep 2023 Max Tegmark, Steve Omohundro

We describe a path to humanity safely thriving with powerful Artificial General Intelligences (AGIs) by building them to provably satisfy human-specified requirements.

Discovering New Interpretable Conservation Laws as Sparse Invariants

1 code implementation31 May 2023 Ziming Liu, Patrick Obin Sturm, Saketh Bharadwaj, Sam Silva, Max Tegmark

Discovering conservation laws for a given dynamical system is important but challenging.

Seeing is Believing: Brain-Inspired Modular Training for Mechanistic Interpretability

1 code implementation4 May 2023 Ziming Liu, Eric Gan, Max Tegmark

We introduce Brain-Inspired Modular Training (BIMT), a method for making neural networks more modular and interpretable.

GenPhys: From Physical Processes to Generative Models

no code implementations5 Apr 2023 Ziming Liu, Di Luo, Yilun Xu, Tommi Jaakkola, Max Tegmark

We introduce a general family, Generative Models from Physical Processes (GenPhys), where we translate partial differential equations (PDEs) describing physical processes to generative models.

The Quantization Model of Neural Scaling

1 code implementation NeurIPS 2023 Eric J. Michaud, Ziming Liu, Uzay Girit, Max Tegmark

We tentatively find that the frequency at which these quanta are used in the training distribution roughly follows a power law corresponding with the empirical scaling exponent for language models, a prediction of our theory.

Language Modelling Quantization

Precision Machine Learning

1 code implementation24 Oct 2022 Eric J. Michaud, Ziming Liu, Max Tegmark

We explore unique considerations involved in fitting ML models to data with very high precision, as is often required for science applications.

Omnigrok: Grokking Beyond Algorithmic Data

1 code implementation3 Oct 2022 Ziming Liu, Eric J. Michaud, Max Tegmark

Grokking, the unusual phenomenon for algorithmic datasets where generalization happens long after overfitting the training data, has remained elusive.

Attribute Representation Learning +1

Poisson Flow Generative Models

1 code implementation22 Sep 2022 Yilun Xu, Ziming Liu, Max Tegmark, Tommi Jaakkola

We interpret the data points as electrical charges on the $z=0$ hyperplane in a space augmented with an additional dimension $z$, generating a high-dimensional electric field (the gradient of the solution to Poisson equation).

Image Generation

AI Poincaré 2.0: Machine Learning Conservation Laws from Differential Equations

no code implementations23 Mar 2022 Ziming Liu, Varun Madhavan, Max Tegmark

We present a machine learning algorithm that discovers conservation laws from differential equations, both numerically (parametrized as neural networks) and symbolically, ensuring their functional independence (a non-linear generalization of linear independence).

BIG-bench Machine Learning

Fault-Tolerant Neural Networks from Biological Error Correction Codes

no code implementations25 Feb 2022 Alexander Zlokapa, Andrew K. Tan, John M. Martyn, Ila R. Fiete, Max Tegmark, Isaac L. Chuang

It has been an open question in deep learning if fault-tolerant computation is possible: can arbitrarily reliable computation be achieved using only unreliable neurons?

Open-Ended Question Answering

Machine-learning hidden symmetries

no code implementations20 Sep 2021 Ziming Liu, Max Tegmark

We present an automated method for finding hidden symmetries, defined as symmetries that become manifest only in a new coordinate system that must be discovered.

BIG-bench Machine Learning

AI Poincaré: Machine Learning Conservation Laws from Trajectories

no code implementations9 Nov 2020 Ziming Liu, Max Tegmark

We present AI Poincar\'e, a machine learning algorithm for auto-discovering conserved quantities using trajectory data from unknown dynamical systems.

BIG-bench Machine Learning Test

AI Feynman 2.0: Pareto-optimal symbolic regression exploiting graph modularity

2 code implementations NeurIPS 2020 Silviu-Marian Udrescu, Andrew Tan, Jiahai Feng, Orisvaldo Neto, Tailin Wu, Max Tegmark

We present an improved method for symbolic regression that seeks to fit data to formulas that are Pareto-optimal, in the sense of having the best accuracy for a given complexity.

regression Symbolic Regression +1

Symbolic Pregression: Discovering Physical Laws from Distorted Video

no code implementations19 May 2020 Silviu-Marian Udrescu, Max Tegmark

We present a method for unsupervised learning of equations of motion for objects in raw and optionally distorted unlabeled video.

regression Symbolic Regression

Foreground modelling via Gaussian process regression: an application to HERA data

no code implementations13 Apr 2020 Abhik Ghosh, Florent Mertens, Gianni Bernardi, Mário G. Santos, Nicholas S. Kern, Christopher L. Carilli, Trienko L. Grobler, Léon V. E. Koopmans, Daniel C. Jacobs, Adrian Liu, Aaron R. Parsons, Miguel F. Morales, James E. Aguirre, Joshua S. Dillon, Bryna J. Hazelton, Oleg M. Smirnov, Bharat K. Gehlot, Siyanda Matika, Paul Alexander, Zaki S. Ali, Adam P. Beardsley, Roshan K. Benefo, Tashalee S. Billings, Judd D. Bowman, Richard F. Bradley, Carina Cheng, Paul M. Chichura, David R. DeBoer, Eloy de Lera Acedo, Aaron Ewall-Wice, Gcobisa Fadana, Nicolas Fagnoni, Austin F. Fortino, Randall Fritz, Steve R. Furlanetto, Samavarti Gallardo, Brian Glendenning, Deepthi Gorthi, Bradley Greig, Jasper Grobbelaar, Jack Hickish, Alec Josaitis, Austin Julius, Amy S. Igarashi, MacCalvin Kariseb, Saul A. Kohn, Matthew Kolopanis, Telalo Lekalake, Anita Loots, David MacMahon, Lourence Malan, Cresshim Malgas, Matthys Maree, Zachary E. Martinot, Nathan Mathison, Eunice Matsetela, Andrei Mesinger, Abraham R. Neben, Bojan Nikolic, Chuneeta D. Nunhokee, Nipanjana Patra, Samantha Pieterse, Nima Razavi-Ghods, Jon Ringuette, James Robnett, Kathryn Rosie, Raddwine Sell, Craig Smith, Angelo Syce, Max Tegmark, Nithyanandan Thyagarajan, Peter K. G. Williams, Haoxuan Zheng

The key challenge in the observation of the redshifted 21-cm signal from cosmic reionization is its separation from the much brighter foreground emission.

Cosmology and Nongalactic Astrophysics

Pareto-optimal data compression for binary classification tasks

1 code implementation23 Aug 2019 Max Tegmark, Tailin Wu

The goal of lossy data compression is to reduce the storage cost of a data set $X$ while retaining as much information as possible about something ($Y$) that you care about.

Binary Classification Classification +5

Learnability for the Information Bottleneck

no code implementations ICLR Workshop LLD 2019 Tailin Wu, Ian Fischer, Isaac L. Chuang, Max Tegmark

However, in practice, not only is $\beta$ chosen empirically without theoretical guidance, there is also a lack of theoretical understanding between $\beta$, learnability, the intrinsic nature of the dataset and model capacity.

Representation Learning

AI Feynman: a Physics-Inspired Method for Symbolic Regression

2 code implementations27 May 2019 Silviu-Marian Udrescu, Max Tegmark

A core challenge for both physics and artificial intellicence (AI) is symbolic regression: finding a symbolic expression that matches data from an unknown function.

regression Symbolic Regression +1

Latent Representations of Dynamical Systems: When Two is Better Than One

no code implementations9 Feb 2019 Max Tegmark

A popular approach for predicting the future of dynamical systems involves mapping them into a lower-dimensional "latent space" where prediction is easier.

Vocal Bursts Valence Prediction

Toward an AI Physicist for Unsupervised Learning

1 code implementation24 Oct 2018 Tailin Wu, Max Tegmark

We investigate opportunities and challenges for improving unsupervised machine learning using four common strategies with a long history in physics: divide-and-conquer, Occam's razor, unification and lifelong learning.


Meta-learning autoencoders for few-shot prediction

1 code implementation26 Jul 2018 Tailin Wu, John Peurifoy, Isaac L. Chuang, Max Tegmark

Compared to humans, machine learning models generally require significantly more training examples and fail to extrapolate from experience to solve previously unseen challenges.


Nanophotonic Particle Simulation and Inverse Design Using Artificial Neural Networks

1 code implementation18 Oct 2017 John Peurifoy, Yichen Shen, Li Jing, Yi Yang, Fidel Cano-Renteria, Brendan Delacy, Max Tegmark, John D. Joannopoulos, Marin Soljacic

We propose a method to use artificial neural networks to approximate light scattering by multilayer nanoparticles.

Computational Physics Applied Physics Optics

Gated Orthogonal Recurrent Units: On Learning to Forget

1 code implementation8 Jun 2017 Li Jing, Caglar Gulcehre, John Peurifoy, Yichen Shen, Max Tegmark, Marin Soljačić, Yoshua Bengio

We present a novel recurrent neural network (RNN) based model that combines the remembering ability of unitary RNNs with the ability of gated RNNs to effectively forget redundant/irrelevant information in its memory.

Ranked #7 on Question Answering on bAbi (Accuracy (trained on 1k) metric)

Denoising Question Answering

The power of deeper networks for expressing natural functions

no code implementations ICLR 2018 David Rolnick, Max Tegmark

It is well-known that neural networks are universal approximators, but that deeper networks tend in practice to be more powerful than shallower ones.

Tunable Efficient Unitary Neural Networks (EUNN) and their application to RNNs

4 code implementations ICML 2017 Li Jing, Yichen Shen, Tena Dubček, John Peurifoy, Scott Skirlo, Yann Lecun, Max Tegmark, Marin Soljačić

Using unitary (instead of general) matrices in artificial neural networks (ANNs) is a promising way to solve the gradient explosion/vanishing problem, as well as to enable ANNs to learn long-term correlations in the data.

Permuted-MNIST Test

Why does deep and cheap learning work so well?

no code implementations29 Aug 2016 Henry W. Lin, Max Tegmark, David Rolnick

We show how the success of deep learning could depend not only on mathematics but also on physics: although well-known mathematical theorems guarantee that neural networks can approximate arbitrary functions well, the class of functions of practical interest can frequently be approximated through "cheap learning" with exponentially fewer parameters than generic ones.

Criticality in Formal Languages and Statistical Physics

no code implementations21 Jun 2016 Henry W. Lin, Max Tegmark

We show that the mutual information between two symbols, as a function of the number of symbols between the two, decays exponentially in any probabilistic regular grammar, but can decay like a power law for a context-free grammar.

An Improved Model of Diffuse Galactic Radio Emission from 10 MHz to 5 THz

3 code implementations16 May 2016 Haoxuan Zheng, Max Tegmark, Joshua S. Dillon, Doyeon A. Kim, Adrian Liu, Abraham Neben, Justin Jonas, Patricia Reich, Wolfgang Reich

We present an improved Global Sky Model (GSM) of diffuse galactic radio emission from 10 MHz to 5 THz, whose uses include foreground modeling for CMB and 21 cm cosmology.

Cosmology and Nongalactic Astrophysics Astrophysics of Galaxies Instrumentation and Methods for Astrophysics

Research Priorities for Robust and Beneficial Artificial Intelligence

no code implementations10 Feb 2016 Stuart Russell, Daniel Dewey, Max Tegmark

Success in the quest for artificial intelligence has the potential to bring unprecedented benefits to humanity, and it is therefore worthwhile to investigate how to maximize these benefits while avoiding potential pitfalls.

Friendly Artificial Intelligence: the Physics Challenge

no code implementations2 Sep 2014 Max Tegmark

Relentless progress in artificial intelligence (AI) is increasingly raising concerns that machines will replace humans on the job market, and perhaps altogether.

A model of diffuse Galactic Radio Emission from 10 MHz to 100 GHz

2 code implementations12 Feb 2008 Angelica de Oliveira-Costa, Max Tegmark, B. M. Gaensler, Justin Jonas, T. L. Landecker, Patricia Reich

Understanding diffuse Galactic radio emission is interesting both in its own right and for minimizing foreground contamination of cosmological measurements.

Optical Character Recognition Optical Character Recognition (OCR)

Separating the Early Universe from the Late Universe: cosmological parameter estimation beyond the black box

1 code implementation2 Jul 2002 Max Tegmark, Matias Zaldarriaga

We present a method for measuring the cosmic matter budget without assumptions about speculative Early Universe physics, and for measuring the primordial power spectrum P*(k) non-parametrically, either by combining CMB and LSS information or by using CMB polarization.


The importance of quantum decoherence in brain processes

no code implementations5 Jul 1999 Max Tegmark

Based on a calculation of neural decoherence rates, we argue that that the degrees of freedom of the human brain that relate to cognitive processes should be thought of as a classical rather than quantum system, i. e., that there is nothing fundamentally wrong with the current classical approach to neural network simulations.

On the dimensionality of spacetime

no code implementations25 Feb 1997 Max Tegmark

Some superstring theories have more than one effective low-energy limit, corresponding to classical spacetimes with different dimensionalities.

General Relativity and Quantum Cosmology High Energy Physics - Phenomenology High Energy Physics - Theory

Cannot find the paper you are looking for? You can Submit a new open access paper.