Polymorphic Combinatorial Frameworks (PCF): Guiding the Design of Mathematically-Grounded, Adaptive AI Agents
David Pearl, Matthew M. Murphy, James Intrilgator · 2025
Read Polymorphic Combinatorial Frameworks (PCF): Guiding the Design of Mathematically-Grounded, Adaptive AI Agents by David Pearl, Matthew M. Murphy, James Intrilgator (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-2508-01581.txt.
Description
David Pearl, Matthew Murphy, and James Intriligator wrote a 2025 paper on a design frame for flexible AI agents. Most agents are built with fixed traits. In fast-changing settings, fixed traits fall behind what the moment demands.
Their frame, called PCF, pairs large language models with formal math. An agent's traits sit in a five-part space named SPARK: skills, personalities, approaches, resources, and knowledge. Points in this space can be reshuffled in real time, so the agent remakes itself as tasks shift. The math base draws on combinatorics, topos theory, and rough fuzzy sets. The authors pitch PCF as a guide for agents that adapt with proof, not guesswork.
Catalog record
| CREATOR | David Pearl, Matthew M. Murphy, James Intrilgator |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-01581