Cubing for Tuning

Cubing for Tuning” by Haoze Wu, Clark W. Barrett, and Nina Narodytska. In Fortieth AAAI Conference on Artificial Intelligence, Thirty-Eighth Conference on Innovative Applications of Artificial Intelligence, Sixteenth Symposium on Educational Advances in Artificial Intelligence, AAAI 2026, Singapore, January 20-27, 2026, (Sven Koenig, Chad Jenkins, and Matthew E. Taylor, eds.), Jan. 2026, pp. 14361-14370.

Abstract

We are exploring the problem of building an automated reasoning procedure that adaptively tunes the high-level solving strategy for a given problem. There are two main distinctive characteristics of our approach: tuning is performed solely online, unlike the common use of tuning as an offline process; and tuning data comes exclusively from the given instance, so we do not rely on the availability of similar benchmarks and can work with unique challenging instances. Our approach builds on top of the divide-and-conquer paradigm that naturally serves partitioned sub-problems for an automated tuning algorithm to obtain a good solving strategy. We demonstrate performance improvement on two classes of important problems-SAT-solving and neural network verification-and show that our method can learn unconventional solving strategies in some cases.

BibTeX entry:

@inproceedings{WBN26,
   author = {Haoze Wu and Clark W. Barrett and Nina Narodytska},
   editor = {Sven Koenig and Chad Jenkins and Matthew E. Taylor},
   title = {Cubing for Tuning},
   booktitle = {Fortieth {AAAI} Conference on Artificial Intelligence,
	Thirty-Eighth Conference on Innovative Applications of Artificial
	Intelligence, Sixteenth Symposium on Educational Advances in
	Artificial Intelligence, {AAAI} 2026, Singapore, January 20-27,
	2026},
   pages = {14361--14370},
   publisher = {{AAAI} Press},
   month = jan,
   year = {2026},
   doi = {10.1609/AAAI.V40I17.38451},
   url = {https://doi.org/10.1609/aaai.v40i17.38451}
}

(This webpage was created with bibtex2web.)