CoEvo: Continual Evolution of Symbolic Solutions Using Large Language Models
Ping Guo, Qingfu Zhang, Xi Lin · 2024
Read CoEvo: Continual Evolution of Symbolic Solutions Using Large Language Models by Ping Guo, Qingfu Zhang, Xi Lin (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-18890.txt.
Description
Ping Guo, Qingfu Zhang, and Xi Lin build a system that grows symbolic finds with language models in a 2024 paper. Symbolic finds are formulas, rules, and code forms at the heart of science. Old search tools find them slow and drop what they learn. New model tools search fast but stall at one pass. The team's CoEvo runs the search as a loop that never quite ends.
A live knowledge bank stores each find and its notes for reuse. Finds wear three forms at once: plain words, math forms, and code. That spread helps the model think in the right tongue for each step. Tests show the loop finds more, faster, than one-pass tools. The authors frame search as a lifelong craft, like human science.
Catalog record
| CREATOR | Ping Guo, Qingfu Zhang, Xi Lin |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 24 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-18890