English
Related papers

Related papers: Toy Models of Superposition

200 papers

Neural networks have greatly boosted performance in computer vision by learning powerful representations of input data. The drawback of end-to-end training for maximal overall performance are black-box models whose hidden representations…

Computer Vision and Pattern Recognition · Computer Science 2020-04-29 Patrick Esser , Robin Rombach , Björn Ommer

We examine the issue of stability of probability in reasoning about complex systems with uncertainty in structure. Normally, propositions are viewed as probability functions on an abstract random graph where it is implicitly assumed that…

Artificial Intelligence · Computer Science 2017-12-14 Subhash Kak

Deep neural networks that yield human interpretable decisions by architectural design have lately become an increasingly popular alternative to post hoc interpretation of traditional black-box models. Among these networks, the arguably most…

Computer Vision and Pattern Recognition · Computer Science 2021-06-24 Adrian Hoffmann , Claudio Fanconi , Rahul Rade , Jonas Kohler

With the advent of highly predictive but opaque deep learning models, it has become more important than ever to understand and explain the predictions of such models. Existing approaches define interpretability as the inverse of complexity…

Neural language models learn word representations that capture rich linguistic and conceptual information. Here we investigate the embeddings learned by neural machine translation models. We show that translation-based embeddings outperform…

Computation and Language · Computer Science 2014-11-14 Felix Hill , KyungHyun Cho , Sebastien Jean , Coline Devin , Yoshua Bengio

Categories of polymorphic lenses in computer science, and of open games in compositional game theory, have a curious structure that is reminiscent of compact closed categories, but differs in some crucial ways. Specifically they have a…

Category Theory · Mathematics 2017-09-19 Jules Hedges

The Platonic Representation Hypothesis suggests that neural networks trained on different modalities (e.g., text and images) align and eventually converge toward the same representation of reality. If true, this has significant implications…

Computer Vision and Pattern Recognition · Computer Science 2026-04-21 A. Sophia Koepke , Daniil Zverev , Shiry Ginosar , Alexei A. Efros

Embedding learning, a.k.a. representation learning, has been shown to be able to model large-scale semantic knowledge graphs. A key concept is a mapping of the knowledge graph to a tensor representation whose entries are predicted by models…

Artificial Intelligence · Computer Science 2016-05-10 Volker Tresp , Cristóbal Esteban , Yinchong Yang , Stephan Baier , Denis Krompaß

'Capsule' models try to explicitly represent the poses of objects, enforcing a linear relationship between an object's pose and that of its constituent parts. This modelling assumption should lead to robustness to viewpoint changes since…

Machine Learning · Computer Science 2021-01-07 Lewis Smith , Lisa Schut , Yarin Gal , Mark van der Wilk

In this paper, we attempt to link the inner workings of a neural language model to linguistic theory, focusing on a complex phenomenon well discussed in formal linguis- tics: (negative) polarity items. We briefly discuss the leading…

Computation and Language · Computer Science 2018-09-03 Jaap Jumelet , Dieuwke Hupkes

The capabilities of large language models (LLMs) have sparked debate over whether such systems just learn an enormous collection of superficial statistics or a set of more coherent and grounded representations that reflect the real world.…

Machine Learning · Computer Science 2024-03-05 Wes Gurnee , Max Tegmark

Word embeddings are ubiquitous in NLP and information retrieval, but it is unclear what they represent when the word is polysemous. Here it is shown that multiple word senses reside in linear superposition within the word embedding and…

Computation and Language · Computer Science 2018-12-10 Sanjeev Arora , Yuanzhi Li , Yingyu Liang , Tengyu Ma , Andrej Risteski

Hypersubstitutions are mappings which map operation symbols to terms. Terms can be visualized by trees. Hypersubstitutions can be extended to mappings defined on sets of trees. The nodes of the trees, describing terms, are labelled by…

Rings and Algebras · Mathematics 2008-12-03 Klaus Denecke , Jorg Koppitz , Slavcho Shtrakov

Interpretability and explainability have gained more and more attention in the field of machine learning as they are crucial when it comes to high-stakes decisions and troubleshooting. Since both provide information about predictors and…

Machine Learning · Computer Science 2024-04-26 Benjamin Leblanc , Pascal Germain

Improving the interpretability of brain decoding approaches is of primary interest in many neuroimaging studies. Despite extensive studies of this type, at present, there is no formal definition for interpretability of brain decoding…

Machine Learning · Statistics 2016-06-21 Seyed Mostafa Kia , Andrea Passerini

Superposition -- when a neural network represents more ``features'' than it has dimensions -- seems to pose a serious challenge to mechanistically interpreting current AI systems. Existing theory work studies \emph{representational}…

Machine Learning · Computer Science 2024-08-13 Kaarel Hänni , Jake Mendel , Dmitry Vaintrob , Lawrence Chan

We present a toy theory that is based on a simple principle: the number of questions about the physical state of a system that are answered must always be equal to the number that are unanswered in a state of maximal knowledge. A wide…

Quantum Physics · Physics 2007-05-23 Robert W. Spekkens

Abstract This project presents a system of neural networks to translate between images and melodies. Autoencoders compress the information in samples to abstract representation. A translation network learns a set of correspondences between…

Computer Vision and Pattern Recognition · Computer Science 2021-12-07 Karl Wienand , Wolfgang M. Heckl

Large language models trained under diverse objectives and architectures have been shown to develop increasingly similar internal representations, an observation formalized as the Platonic Representation Hypothesis. Whether this…

Computation and Language · Computer Science 2026-05-25 Muhammad Usama , Dong Eui Chang

Distributed word embeddings have yielded state-of-the-art performance in many NLP tasks, mainly due to their success in capturing useful semantic information. These representations assign only a single vector to each word whereas a large…

Machine Learning · Computer Science 2020-02-04 Shobhit Jain , Sravan Babu Bodapati , Ramesh Nallapati , Anima Anandkumar
‹ Prev 1 4 5 6 7 8 10 Next ›