#llm4code search results

#ai #LLM #LLM4Code #aircode

In the Agentic AI era, tests are the new specs! 🚀Check out our TENET work on code generation, Pass@1: 69.08% (RepoCod), 81.77% (RepoEval), with +9.5% & +2.2% over top agentic baselines. Paper: arxiv.org/abs/2509.24148 #LLMs #Agents #LLM4Code #CodeGeneration @PurdueCS @cerias

Curious about how tests can guide the entire code generation process, not just validate it afterward? 🚀 Check out TENET, our test-driven LLM agent for repository-level code generation! 🔗 Paper Link: arxiv.org/abs/2509.24148 #LLMs #Agents #LLM4Code #CodeGeneration

Announcing the 3rd workshop on #LLM4Code, co-located with @ICSEconf 2026 in Rio de Janeiro, Brazil 🇧🇷! 🎯We are calling for submissions: 🌟Website: llm4code.github.io 🌟Deadline: Oct 20, 2025 🌟8-page research paper / 4-page position paper (including references) Our…

We introduce CodeARC, a new benchmark for evaluating LLMs’ inductive reasoning. Agents must synthesize functions from I/O examples—no natural language, just reasoning. 📄 arxiv.org/pdf/2503.23145 💻 github.com/Anjiang-Wei/Co… 🌐 anjiang-wei.github.io/CodeARC-Websit… #LLM #Reasoning #LLM4Code #ARC

Two of our papers have been accepted to the #ACL2025 main conference! Try our code generation benchmark 🐟RepoCod (lt-asset.github.io/REPOCOD/) and website generation tool 🧇WAFFLE (github.com/lt-asset/Waffle)! #LLM4Code #MLLM #FrontendDev #WebDev #LLM #CodeGeneration #Security

Excited to be at #ICSE2025 in #Ottawa! Join me for our talk: "Between Lines of Code: Unraveling the Distinct Patterns of Machine and Human Programmers" on 📅 Thursday, May 1st, at 2pm in📌 Room 207. Looking forward to discussing more about this fascinating topic! #LLM4Code #AI

GPT 4.5 ranks only 10th for realistic complex coding tasks from GitHub repositories lt-asset.github.io/REPOCOD/#lite RepoCod tasks are - General code generation tasks, and  - complex tasks: longest average canonical solution length (331.6 tokens) x.com/lin0tan/status… #LLM4Code

(2) democratizing programming to people with little technical background, through AI systems capable of self-planning with easy-to-follow explanations, enabling users to program through natural languages. #LLM4Code #security

🚨Just 2 days left to submit to @llm4code 2025, co-located with ICSE🚀 🎯We welcome papers on various topics, including but not limited to: 🌟LLMs/agents for coding tasks 🌟Code model design/optimization 🌟Datasets/evaluation 🔗Submit now: llm4code.github.io #LLM4Code

Can language models replace developers? RepoCod says “Not Yet”, because GPT-4o and other LLMs have <30% accuracy/pass@1 on real-world method-level code generation tasks. Leaderboard lt-asset.github.io/REPOCOD/ #LLM4code #LLM #CodeGeneration #Security @cerias @PurdueScience

All LLMs including GPT-4o achieve < 30% pass@1 on real-world code completion: Check out 🐟REPOCOD, a real-world code generation benchmark: - Repository-level context - Whole function generation - Validation with test cases - 980 instances from 11 Python projects #LLM #LLM4Code

Wanna generate webpages automatically? Meet🧇WAFFLE: multimodal #LLMs for generating HTML from UI images! It uses structure-aware attention & contrastive learning. arxiv.org/abs/2410.18362 #LLM4Code #MLLM #FrontendDev #WebDev @LiangShanchao @NanJiang719 Shangshu Qian

Introducing Nova, a series of foundation models for binary/assembly code. We have also released fine-tuned models for binary code decompilation. #LLM #LLM4Code #BinaryAnalysis #Security @NanJiang719, Chengxiao Wang, Kevin Liu, @XiangzheX, Xiangyu Zhang, Petr Babkin.

Important Dates: Full, short, position papers and replication studies: - Paper Submission: November 11, 2024 - Author Notification: December 8, 2024 Tool submission: December 12, 2024 WORKSHOP: Sun 27-Mon 28 April 2025, Ottawa, Canada @NLBSE_workshop #NLP #LLM #LLM4Code

#ai #LLM #LLM4Code #aircode

If you missed the fantastic keynotes on #CodeLlama and #Starcoder2 from their initiators and core contributors @b_roziere and @LoubnaBenAllal1, you can still check out the slides tinyurl.com/llm4code24 Hope to see your submissions to the 2nd @LLM4Code workshop. #LLM4code #llm

Introducing Nova, a series of foundation models for binary/assembly code. We have also released fine-tuned models for binary code decompilation. #LLM #LLM4Code #BinaryAnalysis #Security @NanJiang719, Chengxiao Wang, Kevin Liu, @XiangzheX, Xiangyu Zhang, Petr Babkin.

Curious about how tests can guide the entire code generation process, not just validate it afterward? 🚀 Check out TENET, our test-driven LLM agent for repository-level code generation! 🔗 Paper Link: arxiv.org/abs/2509.24148 #LLMs #Agents #LLM4Code #CodeGeneration

Wanna generate webpages automatically? Meet🧇WAFFLE: multimodal #LLMs for generating HTML from UI images! It uses structure-aware attention & contrastive learning. arxiv.org/abs/2410.18362 #LLM4Code #MLLM #FrontendDev #WebDev @LiangShanchao @NanJiang719 Shangshu Qian

All LLMs including GPT-4o achieve < 30% pass@1 on real-world code completion: Check out 🐟REPOCOD, a real-world code generation benchmark: - Repository-level context - Whole function generation - Validation with test cases - 980 instances from 11 Python projects #LLM #LLM4Code

Very excited to share the 1st @llm4code workshop has attracted 170+ registrations! If you’re at @ICSEconf and interested in LLMs, pls join us on April 20, we have 24 presentations and two keynotes on Code Llama (@b_roziere) and StarCoder2 (@LoubnaBenAllal1)! #icse24 #llm4code

We introduce CodeARC, a new benchmark for evaluating LLMs’ inductive reasoning. Agents must synthesize functions from I/O examples—no natural language, just reasoning. 📄 arxiv.org/pdf/2503.23145 💻 github.com/Anjiang-Wei/Co… 🌐 anjiang-wei.github.io/CodeARC-Websit… #LLM #Reasoning #LLM4Code #ARC

Two of our papers have been accepted to the #ACL2025 main conference! Try our code generation benchmark 🐟RepoCod (lt-asset.github.io/REPOCOD/) and website generation tool 🧇WAFFLE (github.com/lt-asset/Waffle)! #LLM4Code #MLLM #FrontendDev #WebDev #LLM #CodeGeneration #Security

In the Agentic AI era, tests are the new specs! 🚀Check out our TENET work on code generation, Pass@1: 69.08% (RepoCod), 81.77% (RepoEval), with +9.5% & +2.2% over top agentic baselines. Paper: arxiv.org/abs/2509.24148 #LLMs #Agents #LLM4Code #CodeGeneration @PurdueCS @cerias

🎉 Excited to announce that our paper "ReSym: Harnessing LLMs to Recover Variable and Data Structure Symbols from Stripped Binaries" has won the Distinguished Paper Award at #CCS2024! I'll present tomorrow afternoon in 9-4 session at @acm_ccs. Hope to see you! #LLM #LLM4Code

GPT 4.5 ranks only 10th for realistic complex coding tasks from GitHub repositories lt-asset.github.io/REPOCOD/#lite RepoCod tasks are - General code generation tasks, and  - complex tasks: longest average canonical solution length (331.6 tokens) x.com/lin0tan/status… #LLM4Code

Our #LLM4Code at #icse24 paper: Semantically Aligned Question and Code Generation for Automated Insight Generation leverages the semantic knowledge of LLMs to generate targeted and insightful questions about data and the corresponding code to answer those questions. (4/7)

@katzy_jonathan presenting our #FORGE24 paper on “An Exploratory Investigation into Code License Infringements in Large Language Model Training Datasets”! Talk to him if U are interested in the data used for training a wide range of #llm4code and their implications! @ConfForge

Last week, I enjoyed a visit to Japan, where I participated in the FM+SE #Shonan seminar & gave a talk about our #LLM4Code studies in @AISE_TUDelft at the FM+SE Summit 2024 Tokyo! Big thanks to Prof. Ahmed Hassan, Jack Jiang & Kamei Yasutaka for organizing this incredible event

@aalkaswan1 presenting our recent @ICSEconf work with @avandeursen on “Traces of Memorisation in Large Language Models for Code” in the final ICSE session on language models and code generation! @AISE_TUDelft @serg_delft #llm #llm4code #ICSE2024

Excited to be at #ICSE2025 in #Ottawa! Join me for our talk: "Between Lines of Code: Unraveling the Distinct Patterns of Machine and Human Programmers" on 📅 Thursday, May 1st, at 2pm in📌 Room 207. Looking forward to discussing more about this fascinating topic! #LLM4Code #AI

🚨 Preprint Alert: Empirical study of the perception and usage of LLMs in an academic software engineering project. Accepted at #LLM4Code workshop co-located with #ICSE 2024. Provides insights into the role of LLMs for SE in an academic setting. arxiv.org/abs/2401.16186

## MoonBit: The AI-Friendly Programming Language of the Future 🌕🤖 **#LLM4Code 2024** will see the unveiling of MoonBit, a revolutionary programming language designed for the AI era! 🤯 **Key features:** * **End-to-end workflow:** Focus on expressing ideas, not tools. AI will…

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