Goto

Collaborating Authors

 mccallum


College student develops LEGO-like, 3D-printed moon bricks

Popular Science

The prototypes made from artificial lunar regolith don't require any binding. More information Adding us as a Preferred Source in Google by using this link indicates that you would like to see more of our content in Google News results. Regolith contains no organic material, and will not naturally bind by itself. A permanent lunar base will require structures resilient enough to withstand everything from cosmic radiation and micro-meteorites to moonquakes . There are plenty of building materials to choose from on Earth, but they all come with a catch: transportation storage aboard spacecraft is limited, and each additional pound of cargo will quickly ramp up costs.


OpenAI Is Pissing Off a Bunch of Mathematicians--Again

WIRED

"There's a perception of mobster behavior" from leading AI companies, one mathematician tells WIRED as OpenAI prepares to release more than 100 new solutions to unsolved problems. In August, OpenAI convened around 40 mathematicians to discuss what to do if AI outpaces human capabilities in the field. The company hinted that its powerful models had solved hundreds of long-standing math problems, say people who were in attendance, but company representatives assured attendees it would not release the solutions all at once--an assurance OpenAI spokesperson Lindsay McCallum says the company is "not aware of"--and, worried by how the community might react, sought advice on how to publish the findings. Attendees responded with "a mixture of excitement and dread," but the meeting was a promising first step, recalls Northwestern University mathematician Bryna Kra. The group, she tells WIRED, asked the firm not to just publish them in a blog or tweet, like they had done with 10 problems earlier that month .





Linear Regression in p-adic metric spaces

arXiv.org Artificial Intelligence

Many real-world machine learning problems involve inherently hierarchical data, yet traditional approaches rely on Euclidean metrics that fail to capture the discrete, branching nature of hierarchical relationships. We present a theoretical foundation for machine learning in p-adic metric spaces, which naturally respect hierarchical structure. Our main result proves that an n-dimensional plane minimizing the p-adic sum of distances to points in a dataset must pass through at least n + 1 of those points -- a striking contrast to Euclidean regression that highlights how p-adic metrics better align with the discrete nature of hierarchical data. As a corollary, a polynomial of degree n constructed to minimise the p-adic sum of residuals will pass through at least n + 1 points. As a further corollary, a polynomial of degree n approximating a higher degree polynomial at a finite number of points will yield a difference polynomial that has distinct rational roots. We demonstrate the practical significance of this result through two applications in natural language processing: analyzing hierarchical taxonomies and modeling grammatical morphology. These results suggest that p-adic metrics may be fundamental to properly handling hierarchical data structures in machine learning. In hierarchical data, interpolation between points often makes less sense than selecting actual observed points as representatives.



Structured Energy Network as a Loss Function Jay-Y oon Lee

Neural Information Processing Systems

Belanger & McCallum (2016) and Gygli et al. (2017) have shown that energy In this work, we propose Structured Energy As Loss (SEAL) to take advantage of the expressivity of energy networks without incurring the high inference cost. This raises a question: Can energy networks be used in a way that is as expressive as SPENs, as efficient at inference as feedforward approaches, and also easy to train?



Answering Compositional Queries with Set-Theoretic Embeddings

arXiv.org Artificial Intelligence

The need to compactly and robustly represent item-attribute relations arises in many important tasks, such as faceted browsing and recommendation systems. A popular machine learning approach for this task denotes that an item has an attribute by a high dot-product between vectors for the item and attribute -- a representation that is not only dense, but also tends to correct noisy and incomplete data. While this method works well for queries retrieving items by a single attribute (such as \emph{movies that are comedies}), we find that vector embeddings do not so accurately support compositional queries (such as movies that are comedies and British but not romances). To address these set-theoretic compositions, this paper proposes to replace vectors with box embeddings, a region-based representation that can be thought of as learnable Venn diagrams. We introduce a new benchmark dataset for compositional queries, and present experiments and analysis providing insights into the behavior of both. We find that, while vector and box embeddings are equally suited to single attribute queries, for compositional queries box embeddings provide substantial advantages over vectors, particularly at the moderate and larger retrieval set sizes that are most useful for users' search and browsing.