It has been more than 20 years since Sergio Segura proposed to exploit atomic sets for preprocessing feature models. Over the holidays, we came up with "A Fast Counting-Free Algorithm for Computing Atomic Sets in Feature Models" Paper: arxiv.org/abs/2501.12490 w/ @Aargon06

Our work "Pick Me: Judging Sample Quality with Binary Decision Diagrams" has been accepted into SPLC 2025'25 🎉. This work is based on the Bachelor's Thesis of my amazing student Aaron Molt (who will be looking for a PhD position soon 😉). Details and preprint below ⬇️ #SPLC25

Aaand the extended version got accepted into #SPLC25 🎉 Title "Improving Deterministic BDD Compilation with Counting-Free Computation of Atomic Sets" Joint work with Aaron Molt who will also present it at the conference. Abstract ⤵️

First talk of Aaron Molt at #SPLC25! Aaron is presenting our work towards enabling efficient deterministic BDD compilation - this time around through eliminating atomic sets during preprocessing.

In his second talk in the same (!) session, Aaron Molt presents our novel metric for comparing 2-wise samples: #pickme - this work is based on his Bachelor's thesis. By the way, Aaron will be looking for a PhD position in 2026, so talk to him! #SPLC25

We were in contention for Best Paper!