Kevin Moran

Assistsnt Professor @UCFComputerSci | SWE & AI Research | Prev Prof @GmuCompSci | @Holy_Cross & @WilliamandMary alum | Director @SageSELab | @UCFCyberSP Prof

[Kevin Moran @kevinmoran]

Are you interested in building LLMs that generate trustworthy🤝🏼, robust💪🏼 and secure🔐 code??

Then join us at LangCode 2027 co-located with NAACL’27!! We hope to bring together a multidisciplinary set of researchers to tackle these problems.

SE friends, check this out and consider attending!

https://t.co/Hbtfv1kdRq

[Kevin Moran @kevinmoran]

Thanks for sharing your thoughts @moarbugs! I agree with a lot of your points here. In particular the “publish sound work” part. As long as there are no obvious flaws affecting the claims of papers, and there is something new, let them be published and let the community judge their impact. However, I think we need more control around reviewing and conflicts than you suggest, and this will probably need to be at least partially automated at the scale we are moving toward.

Not sure what the best way to implement this would be — and you give some good suggestions. But I hope future PC chairs and organizers try something, otherwise it will become unsustainable. However, it’s also tough when there seems to be so much “credit” tied to “traditional” paper acceptances, particularly in academia.

I’m also weary to call the growth of the community a bad thing, full stop. Purely AI-generated garbage is the thing we probably want to control, while allowing the community to grow naturally as the popularity of topics in SE and related disciplines pick up steam is actually a good thing.

[Kevin Moran @kevinmoran]

@krismicinski Yeah this seems like the most reasonable path forward right now.

[UCF Computer Science @UCFComputerSci]

We're putting the spotlight on our talented software engineering faculty! So proud to have Kevin Moran, Aashish Yadavally, Paul Gazzillo, Santu Karmaker and Chen Chen on the team. ✨

[UCF Computer Science @UCFComputerSci]

WFTV recently spoke with the team behind the top computer science senior design project, LABBRA Intersection: AI traffic accident detection. Learn more about their work: https://t.co/LXpwB1qm1X

[UCF Computer Science @UCFComputerSci]

They've done it again: our C3 team is headed to nationals! 🎉

The team placed first at the 2026 Southeast Collegiate Cyber Defense Competition, its ninth first-place finish at the regional competition since 2013: https://t.co/BpGT5xjaAV

[UCF Computer Science @UCFComputerSci]

Our programming team is headed to the ICPC World Finals! 🎉

Thomas Meeks, Benjamin Prins, and Brian Barak ’25 placed 5th at the ICPC North America Championship, earning the chance to compete in Dubai this November: https://t.co/72pGozJIHt

We’re SO proud! #ChargeOn

[ICSE @ICSEconf]

The ICPC MIP Award went to

"A Novel Approach for Estimating Truck Factors", by

Guilherme Avelino, Leonardo Passos, Andre Hora and Marco Tulio Valente.

Link to publication: https://t.co/ON3cfD938f

[UCF Institute of Artificial Intelligence @UCFIAI]

We are now accepting applications for our AI Summer Camp! Apply here today: https://t.co/amXc4MJ9of

[Kevin Moran @kevinmoran]

If you have a recent software testing journal paper to present, come present it in Korea at ICST!!

Deadline is Friday, March 6th!

[ICST Conference @ICSTconf]

📣📣📣ICST 2026: Journal-first track ddl extended to March 6!

Upon receiving multiple requests from the community, we have decided to extend the Journal-First track Track submission deadline to March 6, 2026 (AoE)

https://t.co/js3BgeSyp4

[Kevin Moran @kevinmoran]

@karmake2 Congrats!! 🎉

[Oscar Chaparro @ojcchar]

Second call: 𝐒𝐨𝐟𝐭𝐰𝐚𝐫𝐞 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐬 — help us understand what makes code difficult to comprehend and how AI can support you.
Answering this survey takes 𝐣𝐮𝐬𝐭 10 𝐦𝐢𝐧𝐮𝐭𝐞𝐬: https://t.co/ogEFtghWMu

[Jelani Nelson @minilek]

China went from catching up, to the dominant academic institutions publishing in NeurIPS (top AI conference). Tsinghua topped all universities, and Peking only trails MIT+CMU.

Now is not the time to slash federal research funding, but rather to increase it. This is a Code Red.

[Ziyu Yao @ZiyuYao]

🎉Check out our recent papers accepted to #NeurIPS and #EMNLP on #MechInterp of LLMs (I'm hiring Fall'26 PhDs on this topic)

#NeurIPS2025 Failure by Interference: Language Models Make Balanced Parentheses Errors When Faulty Mechanisms Overshadow Sound Ones
(https://t.co/UHc0WkeCd7 w/ @DakingRai, Sam Miller, @kevinmoran)
My recent favourite! We propose a new perspective of "top-down mechanism decomposition" to understand why LMs fail. Surprisingly, even when LMs fail, we can discover reliable internal mechanisms that can successfully solve the task, and we found that the model fails mostly because the faulty mechanisms overshadow the sound ones! Steering the sound mechanisms solves the problem. We proved the idea on a Code Generation task (balanced parentheses) and found it generalizes to Arithmetic Reasoning.

#EMNLP2025 All for One: LLMs Solve Mental Math at the Last Token With Information Transferred From Other Tokens
(https://t.co/0tKM3a8taY w/ @siddarthpm1 @DakingRai @YilunZhou)
We try to understand how LLMs calculate for a + b - c. Humans solve it following a compositional formulation: a+b first, and -c later. But LLMs do not necessarily do the same. We discover a highly faithful subgraph, where LLMs transfer all information to the last token position and complete all calculations there. Such a sparse and non-compositional subgraph surprisingly generalizes to multiple LLMs.

#EMNLP2025 A Survey on Sparse Autoencoders: Interpreting the Internal Mechanisms of Large Language Models
(https://t.co/jDdhmVO4NA w/ @DuMNCH Ninghao Liu, and their students Dong Shu, Xuansheng Wu, Haiyan Zhao, and my student @DakingRai)
If you've enjoyed our recent survey and #ICML tutorial on Mech Interp (https://t.co/AtFz8XEjYB), this survey will give you more details specifically about SAEs! A must-read;)

#EMNLP2025 Feature Extraction and Steering for Enhanced Chain-of-Thought Reasoning in Language Models
(https://t.co/NRkOckrlmG led by @DuMNCH and his students Zihao Li and Xu Wang)
We extract SAE features representing Verbal Process and Symbolic Process of an LLM's CoT reasoning and perform steering to enhance their effect.

Congrats and thanks to all collaborators!

[Shubhra K. Karmaker ("Santu") @karmake2]

Job Posting: Postdoc Position Available https://t.co/7gbdK8J6fQ

@ReviewAcl @emnlpmeeting @naaclmeeting @aclmeeting #NLP #postdoc #UCF

[Kevin Moran @kevinmoran]

@su_tingsu Congrats!

[UCF Computer Science @UCFComputerSci]

UCF’s new Institute for Artificial Intelligence positions the university as a national leader in AI. Led by Trustee Chair @ucfmshah, it unites 25 faculty and industry partners to drive groundbreaking research: https://t.co/11ZoX01Blw

[UCF @UCF]

AI is reshaping the world. UCF’s new Institute of Artificial Intelligence will help shape the future across disciplines, industries and communities.

Discover more: https://t.co/0eUCbQv4Oe

[Mubarak Shah @ucfmshah]

UCF Launches Institute of Artificial Intelligence to Advance Research, Talent Development Across Disciplines: https://t.co/02ZgPFCoQQ

[Mubarak Shah @ucfmshah]

🚀 Proud moment for @UCF!
8 papers from CRCV and the AI Initiative (Aii) have been accepted to #ICML2025, covering topics from embodied intelligence to graph optimization and LLM alignment.
👇 Check out the full list of contributions from our teams:
🧵