Transfer Learning from Foundational Optimization Embeddings to Unsupervised SAT Representations
Koyena Pal 1,2 Serdar Kadıoğlu 1,3 1 AI Center of Excellence, Fidelity Investments, USA 2 Khoury College of Computer Sciences, Northeastern University, USA 3 Department of Computer Science, Brown University, USA [email protected], [email protected] · 2026
Read Transfer Learning from Foundational Optimization Embeddings to Unsupervised SAT Representations by Koyena Pal 1,2 Serdar Kadıoğlu 1,3 1 AI Center of Excellence, Fidelity Investments, USA 2 Khoury College of Computer Sciences, Northeastern University, USA 3 Department of Computer Science, Brown University, USA [email protected], [email protected] (2026) 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-2604-15448.txt.
Description
Pal and Kadioglu ask whether pre-trained embeddings built for mixed-integer programming can transfer to Boolean satisfiability. Mapping CNF formulas onto the same bipartite constraint-variable graphs lets them reuse the model with no design change or labeled fine-tuning. The vectors capture SAT structure well enough for unsupervised grouping and distribution spotting, a first step toward one shared code for optimization and decision problems.
Catalog record
| CREATOR | Koyena Pal 1,2 Serdar Kadıoğlu 1,3 1 AI Center of Excellence, Fidelity Investments, USA 2 Khoury College of Computer Sciences, Northeastern University, USA 3 Department of Computer Science, Brown University, USA [email protected], [email protected] |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23124 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.15448