Reposted by Jay Hennig
Jay Hennig
@jhennig.bsky.social
1.6K followers 557 following 152 posts
Computational neuroscientist interested in how we learn, and dad to twin boys Asst prof at Baylor College of Medicine www.henniglab.org
Computational neuroscientist interested in how we learn, and dad to twin boys Asst prof at Baylor College of Medicine www.henniglab.org
![To begin, distributions are never normal. For some this seems obvious, hardly worth mentioning, but an aphorism given by Cram´er (1946) and attributed to the mathematician Poincar´e remains relevant: “Everyone believes in the [normal] law of errors, the experimenters because they think it is a mathematical theorem, the mathematicians because they think it is an experimental fact.”](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:kwjddjickma3syepdsfufyvi/bafkreifw25h2nnb3ylgdumcsctlyh6wdbxam7ajlnbwnca7hdxef5q6bo4)
![Overview of the paper.
Title: The Emergent Symbolic Structure of Artificial Neural Networks
Authors: Tom McCoy, Paul Soulos, Tal Linzen, Paul Smolensky
Left: Neural networks encode information in vectors (there is then an image of a vector), yet they excel at tasks long thought to require symbolic structure (there is then an image of a symbolic representation, specifically a syntax tree). How do LLMs do it?
Right: We find that LLM representations can be closely approximated with symbolic structures. This approximation lets us edit the structure of an LLM’s output by editing the structure of its internal representations, as shown. There is then an image of two edits to LLMs. In the first one, the original input is 3 + 6 * 8, with an answer of 51. But if we swap the positions of the 3 and the 6, the output becomes 30. In the second one, the original input is a Python command repeating the list [Z, U] three times, producing [Z, U, Z, U, Z, U]. But if we edit the input in a way that adds a Q at the end of the input, the output becomes [Z, U, Q, Z, U, Q, Z, U, Q].](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:bcsvlv7ubeagj5virdvatadd/bafkreic3zm2rpc4vtqrz2t3y4vcl3t2owp2sxay42vuh5zm5zkg4jhr64e)
![Research [25] shows that people [26] enjoy the feeling of winning games [3,5,19]. Enjoying things can lead to longer life expectancy [2], higher economic productivity [9,22,33-40,302] and improve mental health [9]. Yet many games still include losing players [67,4]. A recent study showed that less than half of the players in a Chess [1] match win on average [42]. We have developed an LLM-driven co-creative [66], user-led [11], mixed-initiative [80] assistant for Chess [2], through which players can redefine the rules of Chess [3] using generative AI prompting such that they are the winner of Chess [4]. A study (n=3) showed that 100\% of people felt better, healthier and more self-actualised after using our tool, showing that this may be the way forward for games and human culture more broadly in the future.](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:h5kgtr755iczaqxomeh2pmh7/bafkreicu2p2o3is64z5k37n222hgahp67ssca6ro3nggbjtpwasig2zzbi)
