Zobrazeno 1 - 10
of 12 989
pro vyhledávání: '"Yona, A."'
Mixture-of-Experts (MoE) models improve the efficiency and scalability of dense language models by routing each token to a small number of experts in each layer. In this paper, we show how an adversary that can arrange for their queries to appear in
Externí odkaz:
http://arxiv.org/abs/2410.22884
Large language models (LLMs) are susceptible to memorizing training data, raising concerns due to the potential extraction of sensitive information. Current methods to measure memorization rates of LLMs, primarily discoverable extraction (Carlini et
Externí odkaz:
http://arxiv.org/abs/2410.19482
Scaling inference compute in large language models (LLMs) through repeated sampling consistently increases the coverage (fraction of problems solved) as the number of samples increases. We conjecture that this observed improvement is partially due to
Externí odkaz:
http://arxiv.org/abs/2410.15466
Autor:
Saez-Mollejo, Jaime, Jirovec, Daniel, Schell, Yona, Kukucka, Josip, Calcaterra, Stefano, Chrastina, Daniel, Isella, Giovanni, Rimbach-Russ, Maximilian, Bosco, Stefano, Katsaros, Georgios
Hole spin qubits are rapidly emerging as the workhorse of semiconducting quantum processors because of their large spin-orbit interaction, enabling fast all-electric operations at low power. However, spin-orbit interaction also causes non-uniformitie
Externí odkaz:
http://arxiv.org/abs/2408.03224
Autor:
Ghalebikesabi, Sahra, Bagdasaryan, Eugene, Yi, Ren, Yona, Itay, Shumailov, Ilia, Pappu, Aneesh, Shi, Chongyang, Weidinger, Laura, Stanforth, Robert, Berrada, Leonard, Kohli, Pushmeet, Huang, Po-Sen, Balle, Borja
Advanced AI assistants combine frontier LLMs and tool access to autonomously perform complex tasks on behalf of users. While the helpfulness of such assistants can increase dramatically with access to user information including emails and documents,
Externí odkaz:
http://arxiv.org/abs/2408.02373
Object detection is a pivotal task in computer vision that has received significant attention in previous years. Nonetheless, the capability of a detector to localise objects out of the training distribution remains unexplored. Whilst recent approach
Externí odkaz:
http://arxiv.org/abs/2407.15763
Autor:
Shumailov, Ilia, Hayes, Jamie, Triantafillou, Eleni, Ortiz-Jimenez, Guillermo, Papernot, Nicolas, Jagielski, Matthew, Yona, Itay, Howard, Heidi, Bagdasaryan, Eugene
Exact unlearning was first introduced as a privacy mechanism that allowed a user to retract their data from machine learning models on request. Shortly after, inexact schemes were proposed to mitigate the impractical costs associated with exact unlea
Externí odkaz:
http://arxiv.org/abs/2407.00106
We posit that large language models (LLMs) should be capable of expressing their intrinsic uncertainty in natural language. For example, if the LLM is equally likely to output two contradicting answers to the same question, then its generated respons
Externí odkaz:
http://arxiv.org/abs/2405.16908
Autor:
Gekhman, Zorik, Yona, Gal, Aharoni, Roee, Eyal, Matan, Feder, Amir, Reichart, Roi, Herzig, Jonathan
When large language models are aligned via supervised fine-tuning, they may encounter new factual information that was not acquired through pre-training. It is often conjectured that this can teach the model the behavior of hallucinating factually in
Externí odkaz:
http://arxiv.org/abs/2405.05904
The Segment Anything Model (SAM) is a deep neural network foundational model designed to perform instance segmentation which has gained significant popularity given its zero-shot segmentation ability. SAM operates by generating masks based on various
Externí odkaz:
http://arxiv.org/abs/2404.12285