Abstract:Equality saturation has become a dominant paradigm for equational program optimization. However, it has never been rigorously compared to another approach to the same problem, even though several exist, the most notable being stochastic search. In this paper, we compare equality saturation to stochastic search over five benchmarks to answer the question: are e-graphs actually good?
Submission history
From: Yihong Zhang [view email]
[v1]
Mon, 18 May 2026 18:27:33 UTC (324 KB)
[v2]
Fri, 22 May 2026 01:08:52 UTC (313 KB)