POPL 2026
Sun 11 - Sat 17 January 2026 Rennes, France

This program is tentative and subject to change.

Sun 11 Jan 2026 16:00 - 16:18 at Horizons - LLMs in Auto-Active Verification

Large Language Models have shown potential for program verification, but progress is hindered by the scarcity of verified code for training. We present ATLAS, an automated pipeline that synthesizes verified programs at scale to address this data bottleneck. ATLAS generates complete Dafny programs with specifications, implementations, and proofs, producing 2.7K verified programs from which we extract over 19K training examples—more than 7 per verified program—by decomposing the synthesis process into multiple specialized tasks. Fine-tuning Qwen 2.5 7B Coder on this dataset produces substantial gains: +23 percentage points on DafnyBench and +50 percentage points on DafnySynthesis. These results demonstrate that synthetic verified code can effectively enhance LLM capabilities for formal verification.

This program is tentative and subject to change.

Sun 11 Jan

Displayed time zone: Brussels, Copenhagen, Madrid, Paris change

16:00 - 18:00
LLMs in Auto-Active VerificationDafny at Horizons
16:00
18m
Talk
ATLAS: Automated Toolkit for Large-Scale Verified Code Synthesis
Dafny
Mantas Bakšys University of Cambridge, Stefan Zetzsche Amazon Web Services, Olivier Bouissou Amazon Web Services, Soonho Kong Amazon Web Services, Remi Delmas Amazon Web Services
Pre-print
16:18
18m
Talk
A benchmark for vericoding: formally verified program synthesis
Dafny
Sergiu Bursuc Beneficial AI Foundation, Theodore Ehrenborg Beneficial AI Foundation, Shaowei Lin Beneficial AI Foundation, Lăcrămioara Aștefănoaei Beneficial AI Foundation, Ionel Emilian Chiosa MIT, Jure Kukovec Beneficial AI Foundation, Alok Singh Beneficial AI Foundation, Oliver Butterley Beneficial AI Foundation, Adem Bizid MIT, Quinn Dougherty Beneficial AI Foundation, Miranda Zhao MIT, Max Tan MIT, Max Tegmark Massachusetts Institute of Technology
Pre-print File Attached
16:36
18m
Talk
DafnyPro: LLM-Assisted Automated Verification for Dafny Programs
Dafny
Debangshu Banerjee UIUC, Olivier Bouissou Amazon Web Services, Stefan Zetzsche Amazon Web Services
Pre-print
16:54
18m
Talk
MiniF2F-Dafny: LLM-Guided Mathematical Theorem Proving via Auto-Active Verification
Dafny
Mantas Bakšys University of Cambridge, Stefan Zetzsche Amazon Web Services, Olivier Bouissou Amazon Web Services
Pre-print
17:12
18m
Talk
Specification-Guided Repair of Arithmetic Errors in Dafny Programs using LLMs
Dafny
Valentina Wu Faculdade de Engenharia, Universidade do Porto, Alexandra Mendes Faculty of Engineering, University of Porto & INESC TEC, Alexandre Abreu University of Porto & INESC TEC
17:30
18m
Talk
Toward Automated, Contamination-free Dafny Benchmark Generation
Dafny
Changjie Wang KTH Royal Institute of Technology, Mariano Scazzariello RISE Research Institutes of Sweden, Dejan Kostic KTH Royal Institute of Technology, Marco Chiesa KTH Royal Institute of Technology
Pre-print
17:48
12m
Day closing
Day closing
Dafny
Stefan Zetzsche Amazon Web Services, Yannick Moy ANSSI
Hide past events