Programming Languages, Formal Methods, and Software Engineering group @IllinoisCS (https://xitter.cf/t.co/61fHOptBEa)
Joined June 2020
- Tweets227
- Following48
- Followers1.1K
- Likes62
π Introducing Structured LLM, a new framework for making large language models more aligned, useful, and efficient.
π Check it out here: structuredllm.com/
[5/n] π‘ Why it matters
Structured LLM enables outputs that are both syntactically valid and semantically sound across multiple domains, including programming, math reasoning, and trustworthy AI.
It opens the door to reliable and trustworthy LLM applications.
[6/n]
Contributors: Sasa Misailovic, Gagandeep Singh, Shubham Ugare, Debangshu Banerjee, Tarun Suresh, Rohan Gumaste, Hangoo Kang, Preston Firestone, Madhav Kanda, Adharsh Kamath
#AI #LLM #StructuredOutput #NLP #MachineLearning #CodeGeneration #TrustworthyAI
1/4
πππ° πππ©ππ« accepted @ ππππ ππππ!
@debangshuban18 , @TarunSures41845 , @shubhamdugare, Sasa Misailovic, @ggn_dp_sngh
introduce CRANE β a decoding algorithm that combines LLM reasoning with formal guarantees.
π arxiv.org/abs/2502.09061
3/4
The paper theoretically proves that constrained decoding can still preserve performance, even when the grammar is augmented with Chain-of-Thought steps.
This bridges free-form reasoning with formal output guarantees.
4/4
On symbolic reasoning benchmarks like GSM-symbolic and FOLIO, CRANE achieves:
π Up to +10% accuracy over prior constrained decoding methods
#ICML2025 #LLM #ConstrainedDecoding #FormalMethods
PL/FM/SE at Illinois retweeted
Looks like my AMS Early Career Notice on proof assistants is published! Check it out here: ams.org/journals/notices/202β¦
PL/FM/SE at Illinois retweeted
#ILLINOIS CS professor Darko Marinov & former students Qingzhou Luo, Farah Hariri & Lamyaa Eloussi were awarded 2024 Test of Time Honorable Mention by the ACM ICF of Software Engineering for their 2014 paper "An empirical analysis of flaky tests." πΈπΉ
siebelschool.illinois.edu/neβ¦
ALT Darko Marinov
PL/FM/SE at Illinois retweeted
Yay, our paper on the abilities of transformer models to learn to emulate structural recursion was accepted to TMLR, a wonderful machine learning research journal! With Dylan Zhang, Curt Tigges, Zory Zhang, Stella Biderman, and Maxim Raginsky. Will update the arXiv version soon!
ALT Hi Talia Ringer, This is to inform you that your submitted camera ready version of your paper 2200: Transformer-Based Models Are Not Yet Perfect At Learning to Emulate Structural Recursion has been verified and confirmed by the Action Editor. We thank you again for your contribution to TMLR and congratulate you for your successful submission! The TMLR Editors-in-Chief
PL/FM/SE at Illinois retweeted
Our paper "Correctly Compiling Proofs About Programs Without Proving Compilers Correct" by Audrey Seo*, (@audreyleeseo), Christopher Lam* (@gracelessfaller), Dan Grossman (@djg98115), and Talia Ringer will appear in ITP 2024! Preprint: dependenttyp.es/pdf/potpie.pβ¦
* co-first authors
Our alumni are awesome! Congratulations, @adithyamuralism. What a big win for Wisconsin π
@plfmse students rock π
This Post is from an account that no longer exists. Learn more
PL/FM/SE at Illinois retweeted
@yinlin_deng is presenting her FuzzGPT work at #icse24. FuzzGPT aims to guide LLMs (which were trained to generate more natural and common info) to generate exceptional and edge-case tests for more effective fuzzing.
πarxiv.org/abs/2304.02014
PL/FM/SE at Illinois retweeted
@steven_xia_ is presenting fuzz4all, the first universal fuzzer via large language models, at #icse24! Joint work with Matteo Paltenghi, Jiale Tian, and Michael Pradel.
π§βπ» fuzz4all.github.io
π arxiv.org/abs/2308.04748
