Zobrazeno 1 - 4
of 4
pro vyhledávání: '"mechanistic intepretability"'
Autor:
Kantamneni, Subhash1 (AUTHOR) subhashk@mit.edu, Liu, Ziming1 (AUTHOR), Tegmark, Max1 (AUTHOR)
Publikováno v:
Entropy. Nov2024, Vol. 26 Issue 11, p997. 21p.
Publikováno v:
Entropy, Vol 26, Iss 11, p 997 (2024)
How do transformers model physics? Do transformers model systems with interpretable analytical solutions or do they create an “alien physics” that is difficult for humans to decipher? We have taken a step towards demystifying this larger puzzle b
Externí odkaz:
https://doaj.org/article/de124c4a41a44e98ab0f236da00b2563
The profound success of transformer-based language models can largely be attributed to their ability to integrate relevant contextual information from an input sequence in order to generate a response or complete a task. However, we know very little
Externí odkaz:
http://arxiv.org/abs/2410.02102
Publikováno v:
Nature Communications; 10/30/2024, Vol. 15 Issue 1, p1-18, 18p