Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
AI Blogs
Irrespective of the specific task at hand, there are common safety constraints that we expect our agents to adhere to in any given environment. These constraints ensure that the agents do not engage in actions that may cause harm or…
arXivLabs is a platform where collaborators can create and share new features for arXiv directly on our website. Both individuals and organizations who work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy.…
In this study, we investigate a Geometric Deep Learning model from a thermodynamic perspective, considering the weights as particles that are neither quantum nor relativistic. We analyze the concept of temperature, as described in a previous publication [7], and explore…
[Submitted on 1 Sep 2023] Download a PDF of the paper titled “Jointly Exploring Client Drift and Catastrophic Forgetting in Dynamic Learning” by Niklas Babendererde and 4 other authors Download PDF Abstract: Federated and Continual Learning have emerged as potential…
Introducing a new distributed computing framework that is resilient to slow compute nodes and capable of both approximate and exact linear operations. This innovative approach combines randomized sketching and polar codes within the context of coded computation. We have developed…
[Submitted on 1 Sep 2023] Click here to download a PDF of the paper titled “Taken out of context: On measuring situational awareness in LLMs,” written by Lukas Berglund and 7 other authors: Download PDF Abstract: The purpose of this…
This work presents enhancements to the stability and applicability of Cyclic DARTS (CDARTS), which is a Differentiable Architecture Search (DARTS) based method for neural architecture search (NAS). CDARTS employs a cyclic feedback mechanism to concurrently train the search and evaluation…
The content describes the use of black-box optimization in finding optimal parameters for systems like Neural Networks or complex simulations. It introduces Polynomial-Model-Based Optimization (PMBO) as a novel black-box optimizer that fits a polynomial surrogate to the objective function to…
We focus on studying the learning of $\\epsilon$-optimal strategies in zero-sum imperfect information games (IIG) with trajectory feedback. In this scenario, players update their policies sequentially based on their observations throughout a fixed number of episodes, denoted as $T$. However,…
In today’s era of data-driven technology, there is a growing interest in synthetic data. This refers to artificially generated data that resembles real-world data but does not contain any personal information. The rise in popularity of synthetic data is driven…
