Zobrazeno 1 - 10
of 2 249
pro vyhledávání: '"A A Burtsev"'
This paper addresses the challenge of creating a neural architecture for very long sequences that requires constant time for processing new information at each time step. Our approach, Associative Recurrent Memory Transformer (ARMT), is based on tran
Externí odkaz:
http://arxiv.org/abs/2407.04841
Autor:
Anokhin, Petr, Semenov, Nikita, Sorokin, Artyom, Evseev, Dmitry, Burtsev, Mikhail, Burnaev, Evgeny
Advancements in the capabilities of Large Language Models (LLMs) have created a promising foundation for developing autonomous agents. With the right tools, these agents could learn to solve tasks in new environments by accumulating and updating thei
Externí odkaz:
http://arxiv.org/abs/2407.04363
Autor:
Sagirova, Alsu, Burtsev, Mikhail
Publikováno v:
Cognitive Systems Research, Volume 75, 2022, Pages 16-24, ISSN 1389-0417
Even though Transformers are extensively used for Natural Language Processing tasks, especially for machine translation, they lack an explicit memory to store key concepts of processed texts. This paper explores the properties of the content of symbo
Externí odkaz:
http://arxiv.org/abs/2406.14213
Autor:
Kuratov, Yuri, Bulatov, Aydar, Anokhin, Petr, Rodkin, Ivan, Sorokin, Dmitry, Sorokin, Artyom, Burtsev, Mikhail
In recent years, the input context sizes of large language models (LLMs) have increased dramatically. However, existing evaluation methods have not kept pace, failing to comprehensively assess the efficiency of models in handling long contexts. To br
Externí odkaz:
http://arxiv.org/abs/2406.10149
Autor:
Kuratov, Yuri, Bulatov, Aydar, Anokhin, Petr, Sorokin, Dmitry, Sorokin, Artyom, Burtsev, Mikhail
This paper addresses the challenge of processing long documents using generative transformer models. To evaluate different approaches, we introduce BABILong, a new benchmark designed to assess model capabilities in extracting and processing distribut
Externí odkaz:
http://arxiv.org/abs/2402.10790
Publikováno v:
AIP Advances 14, 015107 (2024)
Various techniques are available in order to obtain information on samples of a different nature in near-field scanning microwave microscopy (NSMM), with transmission-line resonator (TLR) techniques considered as the most advanced in terms of sensiti
Externí odkaz:
http://arxiv.org/abs/2401.05801
Real-world Knowledge Graphs (KGs) often suffer from incompleteness, which limits their potential performance. Knowledge Graph Completion (KGC) techniques aim to address this issue. However, traditional KGC methods are computationally intensive and im
Externí odkaz:
http://arxiv.org/abs/2311.01326
Publikováno v:
Vestnik Permskogo Universiteta: Seriâ Geologiâ, Vol 23, Iss 3, Pp 201-213 (2024)
The 40Ar/39Ar method determined the age of sanidine from high-potassium trachyte fragments of pipe fluid-explosive breccia. Fluid-explosive breccia breaks through the basalts of the Early Devonian Kanino-Timan complex. The established age of sanidine
Externí odkaz:
https://doaj.org/article/3b2d11f9542f4904b0118ab1a376166a
Autor:
Neal, Jacob, Burtsev, Anton, Ribeiro, Jean Helder Marques, Taira, Kunihiko, Theofilis, Vassilios, Amitay, Michael
Experimental investigations were performed to elucidate the features of flow fields occurring over cantilevered finite-aspect ratio NACA 0015 wings at high angles of attack with various sweep angles and taper ratios. Volumetric Stereoscopic Particle
Externí odkaz:
http://arxiv.org/abs/2308.12442
Autor:
Karpov, Dmitry, Burtsev, Mikhail
This article investigates the knowledge transfer from the RuQTopics dataset. This Russian topical dataset combines a large sample number (361,560 single-label, 170,930 multi-label) with extensive class coverage (76 classes). We have prepared this dat
Externí odkaz:
http://arxiv.org/abs/2306.07797