Tessera
Folgen

Seongmin Lee

@nimgnoeseel.bsky.social

🎲 A board game geek, 🏸 a badminton lover, and a postdoc researcher @UCLA working on dynamic and statistical program analysis.

4 Posts18 Follower50 Folgt
Yannic Noller@yannicnoller.bsky.social · last yr.

🚨 Our amazing #FUZZING'25 keynotes are online! "Constraining Fuzzing without Paying Too Much" by Miryung Kim youtu.be/L90MBb6NLBE "Are you sure you belong in academia?" by Will Wilson youtu.be/qQGuQ_4V6WI // @mboehme.bsky.social, László Szekeres, @rohan.padhye.org, @ruijiemeng.bsky.social

Marcel Böhme@mboehme.bsky.social · last yr.

We had two exciting keynotes: * From academia: Miryung Kim (Prof @ UCLA) * From industry: Will Wilson (CEO and Co-Founder of @AntithesisHQ.bsky.social). Stay tuned for recordings!

Seongmin Lee@nimgnoeseel.bsky.social · last yr.

💥 Fuzzing found a crash—but how do we debug it? We present DiffMin, a technique that makes crashing inputs closer to passing ones while keeping the crash. This helps localize and diagnose crashes more precisely. Just accepted to FSE’25 Posters 🎉 📄 arxiv.org/abs/2505.02305 #FSE25 #fuzzing #debugging

Refining Fuzzed Crashing Inputs for Better Fault Diagnosis

We present DiffMin, a technique that refines a fuzzed crashing input to gain greater similarities to given passing inputs to help developers analyze the crashing input to identify the failure-inducing...

arxiv.org

Andreas Zeller@andreaszeller.bsky.social · last yr.

In today's #AST2025 keynote on our new #Fandango fuzzer, I presented ongoing extensions for protocol fuzzing, oracle checking, coverage guidance, much more. Slides now available: conf.researchr.org/details/ast-...

Personalized Fuzzing (AST 2025) - AST 2025

The 6th ACM/IEEE International Conference on Automation of Software Test (AST 2025) Software pervasiveness in both industry and digital society, as well as the proliferation of Artificial Intelligenc...

conf.researchr.org

Andreas Zeller@andreaszeller.bsky.social · last yr.

What do you get if you combine #grammars, #constraints, #evolutionary algorithms, and #Python in one? A mighty fuzzer! Check out our latest #FANDANGO work, to appear at #ISSTA2025: publications.cispa.de/articles/sta... To try out Fandango yourself, check out its home page: fandango-fuzzer.github.io

FANDANGO: Evolving Language-Based Testing José Antonio Zamudio Amaya, Marius Smytzek, Andreas Zeller Language-based fuzzers leverage formal input specifications (languages) to generate arbitrarily large and diverse sets of valid inputs for a program under test. Modern language-based test generators combine grammars and constraints to satisfy syntactic and semantic input constraints. ISLa, the leading input generator in that space, uses symbolic constraint solving to solve input constraints. Using solvers places ISLa among the most precise fuzzers but also makes it slow. In this paper, we explore search-based testing as an alternative to symbolic constraint solving. We employ a genetic algorithm that iteratively generates candidate inputs from an input specification, evaluates them against defined constraints, evolving a population of inputs through syntactically valid mutations and retaining those with superior fitness until the semantic input constraints are met. This evolutionary procedure, analogous to natural genetic evolution, leads to progressively improved inputs that cover both semantics and syntax. This change boosts the efficiency of language-based testing: In our experiments, compared to ISLa, our search-based FANDANGO prototype is faster by one to three orders of magnitude without sacrificing precision. The search-based approach no longer restricts constraints to constraint solvers' (miniature) languages. In FANDANGO, constraints can use the whole Python language and library. This expressiveness gives testers unprecedented flexibility in shaping test inputs. It allows them to state arbitrary goals for test generation: "Please produce 1,000 valid test inputs where the <voltage> field follows a Gaussian distribution but never exceeds 20 mV."
Seongmin Lee@nimgnoeseel.bsky.social · last yr.

Can we foresee the unseen in software testing? I’ll share how statistical magic helps us quantify residual risk and make testing results more accountable and reliable✨Thanks for the invite! 🙏 Join my tutorial @ SBST (co-located w/ ICSE) 🖥️ sbft25.github.io 📅 28.04. 14:30 GMT-4 (live on Twitch too)

Bild
Seongmin Lee@nimgnoeseel.bsky.social · 2y ago

🎉 Excited to share that my 1st paper at @iclr-conf.bsky.social has been accepted! Big thanks to my awesome colleague @mboehme.bsky.social and the fantastic SoftSec group #MPI. Marcel has shared a great post with a brief explanation of our work—check it out! 🚀 I’ll share a public version soon!

Marcel Böhme@mboehme.bsky.social · 2y ago

Just got our first ICLR paper accepted! 🥳🥳 It shines light on a beautiful statistical riddle! Suppose, you train a classifier. How representative was the training data? What is the proportion of the *unknown* population of classifier inputs with classes *not* in the training data? Led by Seongmin.