LLM Agents Implement an NLG System from Scratch: Building Interpretable Rule-Based RDF-to-Text Generators
Mateusz Lango, Ondřej Dušek · 2025
Read LLM Agents Implement an NLG System from Scratch: Building Interpretable Rule-Based RDF-to-Text Generators by Mateusz Lango, Ondřej Dušek (2025) 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-2512-18360.txt.
Description
Mateusz Lango and Ondřej Dušek let a team of AI agents write a text generator from scratch. The job is turning data triples, the linked facts behind knowledge graphs, into plain sentences. Instead of training a network on human examples, their agents draft rule based Python code, test it, and refine it through talk among themselves.
The finished generator is open to read, needs no labeled training texts, and writes in a blink on a single basic processor. On two standard data sets its output made up fewer false facts than tuned or prompted language models, at a small cost in flow. The work shows agent teamwork can yield clear, old school software.
Catalog record
| CREATOR | Mateusz Lango, Ondřej Dušek |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 44 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-18360