
AISE-TUDelft
@aise_tudelft
Artificial Intelligence-based Software Engineering research lab with a focus on addressing challenges of large language models for code. #LLM @TUDelft
ID: 1664570130376867840
02-06-2023 09:50:44
54 Tweet
43 Followers
23 Following

#sensymposium24 presents ‘Evaluating Large Language Models for Code Generation: a Practical View,’ an invited talk by Maliheh Izadi Maliheh (Mali) Izadi from SERG TU Delft. Join us on 7 June at CWI Amsterdam. Bin Lin Casper Bach Tijs van der Storm Jurgen Vinju CWI VERSEN


Yesterday, I arrived in Brazil for a week of exciting events at FSE 2025 2024! 🇧🇷 Here’s a snapshot of the AISE-TUDelft team's schedule, If you are passionate about the #AI4SE domain, we would love to chat with you! 👇 1\n SERG TU Delft


FSE 2025 AISE-TUDelft SERG TU Delft Tuesday, 11:30 AM: Don't miss Aral's presentation on our award-winning paper "A Transformer-Based Approach for Smart Invocation of Automatic Code Completion" at the AIware 2025. We’re delighted to have received the *ACM Distinguished Paper Award* for this work. 🏆🥇


Very insightful keynote by Baishakhi Ray on semantic awareness for code models at AIware 2025 ! “Improve trust, integrate developer feedback, and increase explainability.” #AIWare #FSE2024 #LLM4Code


“Leveraging Large Language Models for Enhancing the Understandability of Generated Unit Tests” accepted for ICSE #ICSE2025 🥳 Pre-print to follow Here already a shout-out to Amirhossein Deljouyi Roham Koohestani Maliheh (Mali) Izadi for the great effort & nice collaboration cc SERG TU Delft


Our recent study accepted in the first cycle of ICSE ! Kudos to Amirhossein Deljouyi & Roham Koohestani for the great job and thanks to Andy Zaidman for the inspiring collaboration 😊 Stay tuend for the preprint! Canada, here we come 🇨🇦 #ICSE #ICSE2025 #softwaretesting #LLM #LLMs4Code

Aral presenting our work on the use of transformers as filters for automatic code completions in IDE. 🏆🥇 AIware 2025 #AIWare colocated with FSE 2025’24 A Transformer-Based Approach for Smart Invocation of Automatic Code Completion: dl.acm.org/doi/pdf/10.114… AISE-TUDelft


Great industry statement talk at AIware 2025 by Satish Chandra. Lots of good statements, one that i fully support is that evals matter & we need better benchmarks! An effort in this direction is our collab w/ JetBrains;Long Code Arena: arxiv.org/pdf/2406.11612 By Egor Bogomolov et al.



This year, AISE-TUDelft wrapped up the academic year with a retreat at the karting track!🏎️💨 We also bid farewell to Fabio Salern, our visiting MSc student who spent the last 5 months w/ us. Your presence was greatly appreciated & we wish you all the best! 1/n #HappySupervisor


🇨🇦 Planning to attend IEEE SANER'25 in Montreal? 🇨🇦 Consider submitting a work to one of our 6 accepted #workshops! They cover various topics including fairness, continuous education with AI/SE, MSR for privacy/security, blockchain, quality assurance, & quantum software analysis.


Last week, I had the pleasure of delivering an invited #talk at the #BITS&CHIPS event where I shared some of our recent work on #evaluating #LLMs4code. A great day filled with insightful discussions & networking opportunities.Huge thanks to Nieke Roos for his excellent efforts!



Today Maliheh (Mali) Izadi has visited us in Eindhoven to discuss the amazing work she is doing on evaluating LLMs on software engineering tasks.



1/6 🎉 Thrilled to announce that our collaborative paper, "Code Red! 🚨 On the Harmfulness of Applying Off-the-shelf Large Language Models to Programming Tasks," has been accepted to FSE 2025 💻 Arie van Deursen Maliheh (Mali) Izadi SERG TU Delft AISE-TUDelft