📢 I’m looking to hire a postdoc to work closely with me and my research group at UT Austin on exciting topics in core PL/FM, as well as applications of PL/FM ideas to other areas. If you are interested, or know someone who might be a great fit, please DM me!
AbsEvolve also handles non-linear operators like quadratic assignments. Check out our paper for more details on this and other supported domains and operators, formal guarantees, and full evaluation: https://t.co/JARwgZkcjk
[N/N]
Excited to share that our paper titled "Evolving Abstract Transformers for Gradient-Guided, Adaptable Abstract Interpretation" has been accepted at PLDI 2026! #PLDI2026
Huge thanks to my collaborators @debangshuban18 and @ggn_dp_sngh!
Details in 🧵
[1/N]
This enables efficient and adaptable analysis across domains: the blue line and bars show how more invariants get strengthened as gradient steps R increase, while the red line shows convergence to the most precise invariants computed by an LP solver baseline, 3.2× faster.
[4/N]
Really nice work! Leveraging LLM’s general ability to reason about code execution and and then generating reusable rewrite rules to make the process reliable by grounding it with traditional compiler techniques; best of both worlds and the way to go!
There’s a lot of discussion around using LLMs to generate compiler code via prompting, but this on-the-fly approach can be unreliable and brittle.
We propose 𝐑𝐮𝐥𝐞𝐅𝐥𝐨𝐰, a more reliable way to use LLMs for optimizing data-analysis programs (e.g, Pandas).
(1/5)
@syhw Neural abstract interpretation is indeed an exciting direction, some work from our group that tries to build small neural networks that serve as abstract interpreters. LLMs are also possible https://t.co/cq8UC0HWiL
Excited to finally share our #NeurIPS2025 paper "🔮PurpCode: Reasoning for Safer Code Generation"! 🙌
👐 First post-training recipe for training safe code reasoning models
🚀 SOTA for cybersafety + utility, outperforming Sonnet 4, o4-mini, R1
🥇 Winner of 2025 Amazon Nova AI Challenge
📝 Paper: https://t.co/9aOoiL5zgJ
🧵👇 1/11
🚀 Introducing Structured LLM, a new framework for making large language models more aligned, useful, and efficient.
👉 Check it out here: https://t.co/Qk9hjpnmVL
Excited and honored to share that my MS Thesis on "Neural Abstract Interpretation" has been awarded the 2024 David J. Kuck Outstanding Master’s Thesis Award at UIUC 🏆🎉!
Thesis Link: https://t.co/5pCOpcVgoH
Nice start to the year🤞
🚀 Excited to present our paper "Relational DNN Verification Leaps Forward With RABBit" at #NeurIPS2024 on December 11th!
Authored by @TarunSures41845 and @debangshuban18!
1/N