Embers of autoregression show how large language models are shaped by the problem they are trained to solve
McCoy, R Thomas, Yao, Shunyu, Friedman, Dan, Hardy, Mathew D, Griffiths, Thomas L
Published in Proceedings of the National Academy of Sciences - PNAS (08.10.2024)
Published in Proceedings of the National Academy of Sciences - PNAS (08.10.2024)
Get full text
Journal Article
Resampling reduces bias amplification in experimental social networks
Hardy, Mathew D, Thompson, Bill D, Krafft, P M, Griffiths, Thomas L
Published in Nature human behaviour (01.12.2023)
Published in Nature human behaviour (01.12.2023)
Get more information
Journal Article
When a language model is optimized for reasoning, does it still show embers of autoregression? An analysis of OpenAI o1
McCoy, R. Thomas, Yao, Shunyu, Friedman, Dan, Hardy, Mathew D, Griffiths, Thomas L
Year of Publication 02.10.2024
Year of Publication 02.10.2024
Get full text
Journal Article
Bias amplification in experimental social networks is reduced by resampling
Hardy, Mathew D, Thompson, Bill D, Krafft, P M, Griffiths, Thomas L
Published in arXiv.org (05.10.2022)
Get full text
Published in arXiv.org (05.10.2022)
Paper