What a proof guarantees depends on what we ask it to prove. Our audit of AI-generated memory-safety proofs explains why specifications matter, using a checked example and findings from our audit of the Anthropic–Theorem verification effort. Read the paper and supporting evidence.
NUS Computing features Umang’s research on data race detection, following the publication of Dynamic Race Detection with O(1) Samples in Communications of the ACM Research Highlights.
NUS Computing features Umang’s 2026 Robert Brown Promising Researcher Award, recognising his research on software verification.
Umang Mathur and Krishna S. (IIT Bombay) are organizing Frontiers in Verification and Analysis for Concurrency, an ACM India IndiCS seminar at the Infosys Mysore campus, 30 November–4 December 2026. The seminar will bring researchers together for talks, open-problem sessions, and discussions on testing, analysis, and verification of concurrent systems. Seminar website.
Our work on automated C-to-Rust translation—including ACToR, Reboot, and program skeletons—is cited in Ben Livshits’s blog post Automating C to Safe Rust Translation.
Instrumentation Optimization for Practical Dynamic Race Detection accepted to appear at ATC 2026.
Ankit Jayant joins us as a Research Assistant. Welcome, Ankit!
Dynamic Race Detection with O(1) Samples accepted to appear in Communications of the ACM Research Highlights.
Awarded a Ministry of Education Tier 2 grant for “Foundations for Testing and Verifying Concurrency in Go”. Research support.