English

Posts with #ML4Code

#ML4Code

Public posts on X that use #ML4Code.

Related hashtags

  1. Anmol Agarwal@anmolagarwal999
    Excited to be in Vienna for @icmlconf to present “Class-Level Code Generation from Natural Language Using Iterative, Tool-Enhanced Reasoning over Repository” ( at @DMLRWorkshop 4PM today. Excited to connect with #ML4Code & LLM Agent enthusiasts! #icml
    19263.3K
  2. Amir M. Mir✓@amir_mir93
    Happy to share that our research paper, "An Empirical Study of Large Language Models for Type and Call Graph Analysis in Python and JavaScript" has been accepted for publication in Empirical Software Engineering (EMSE) journal. #LLM #ML4Code #EMSE #AI
    205193
  3. Justin ("Goju") Gottschlich@j_gottschlich
    I had the honor of giving a keynote at LADSIOS, co-located with VLDB '21, chaired by MIT Prof. @tim_kraska, Harvard Prof. Stratos Idreos, and Microsoft Principal Scientist Umar Farooq Minhas. Love to hear your thoughts! :) #ml4code #ml #ai #mp
    146
  5. Ajay Jain✓@ajayj_
    New paper on contrastive learning for programming languages with @parasjain @tianjun_zhang @pabbeel @mejoeyg Ion Stoica! ContraCode learns similar representations for equivalent but differently implemented JavaScript programs using compiler-based data augmentations #ml4code
    Paras Jain @parasjain
    Can contrastive learning be applied to text? Yes! We propose Contrastive Code Representation Learning: an unsupervised pretext task to learn representations of program *functionality*. Paper w/ @ajayj_ Tianjun Zhang @pabbeel @mejoeyg Ion Stoica 👇 1/7
    0110
  6. Paras Jain✓@parasjain
    Can contrastive learning be applied to text? Yes! We propose Contrastive Code Representation Learning: an unsupervised pretext task to learn representations of program *functionality*. Paper w/ @ajayj_ Tianjun Zhang @pabbeel @mejoeyg Ion Stoica 👇 1/7
    149144
  7. Miltos Allamanis 🇪🇺@miltos1
    Interested in doing a PhD at the intersection of Machine Learning, Programming Languages and Software Engineering? Michael Pradel's excellent group at Stuttgart is looking for candidates: #MLonCode #ML4Code
    12741
  8. AIPLANS Workshop@createaiplans
    Leading up to #AIPLANS, the #ML4Code Montreal RG will host a talk from @miltos1, who will share his insights on Bug Detection & Repair with Machine Learning tomorrow afternoon at 12pm EST. The RG invites AIPLANS attendees to participate in this discussion!
    003
  9. Moderne, Inc.✓@moderneinc
    What happens when Moderne’s fast, 100% accurate code refactoring automation meets #generativeAI? 🚀 Check out this blog from our CTO to learn what it takes to auto-remediate code at scale—and how #ML4Code 🤖can make this even better.
    058794
  10. Paras Jain✓@parasjain
    @parasjain
    We produce semantically equivalent but textually divergent copies of source functions via automated source-to-source compiler transforms. These augmentations delete dead code, rename variables, evaluate arithmetic etc but preserve functionality. 4/7
    Download
    117
  11. Paras Jain✓@parasjain
    @parasjain
    The model is pre-trained using 1.8m JavaScript programs mined from Github @miltos1 @mmjb86. Our work extends the Momentum Contrast approach to code @inkynumbers. Unsupervised learning is a powerful method to scale #ml4code methods. 5/7
    115
  12. Miltos Allamanis 🇪🇺@miltos1
    Recently, we noticed that near-duplicates in code corpora have a negative effect on real-life applications of #ML4Code. I spent some time documenting and understanding this: Many applications are affected (includes ideas published in ML, NLP, PL, SE).
    2827
  13. CodistAI@AiCodist
    Thanks @philipvollet for sharing over 🤩 AI-powered docstring is now available for your Jupyter Notebook scripts! #ml4code #codequality
    Philip Vollet @philipvollet
    Lazy programmer? Guilty when it comes to proper comments? Focus on coding! Docly will handle the comments now also in #Jupyter notebooks The same command, the same ease, more code covered!   Sign up for the beta @AiCodist 🤞
    024
  14. Philip Vollet✓@philipvollet
    Lazy programmer? Guilty when it comes to proper comments? Focus on coding! Docly will handle the comments now also in #Jupyter notebooks The same command, the same ease, more code covered!   Sign up for the beta @AiCodist 🤞
    Download
    0518
  15. Breandan Considine✓@breandan
    Prediction: AB will acquire JB within five years. Both working on #ML4Code. Both invested in dev tools and software teamwork. Language and culture fit. Strategic way to offset MS/GH/OAI monopoly. Hope I'm wrong.
    Breandan Considine @breandan
    Prediction: MS will acquire GH within five years. If the #ML4Code stuff delivers for MS, acquisition is highly likely. Although it would have been cheaper a few years ago.
    405
  16. Breandan Considine✓@breandan
    Prediction: MS will acquire GH within five years. If the #ML4Code stuff delivers for MS, acquisition is highly likely. Although it would have been cheaper a few years ago.
    002
  17. Ajay Jain✓@ajayj_
    What if your IDE crashed with specific variable names? 😱 #ML4Code is powerful, but models like BERT are sensitive to adversarial code edits, performing worse than random guessing! We find that contrastive learning is more robust. Check out ContraCode: 👇
    Download
    12086
  18. Nadia Chirkova@nadiinchi
    Yesterday I defended my PhD thesis on neural networks for source code processing at @CS_HSE! Huge thanks to my advisor, Dmitry Vetrov, for his valuable guidance, to the committee for insightful questions and to @bayesgroup for friendly support during these 4 years! #PhD #ML4Code
    Download
    7286
  19. Bieber Babble@BieberBabble
    Full thread on "A Library for Representing Python Programs as Graphs for Machine Learning" -- read on for details! 🧵 Report: GitHub: PyPi: Install: pip install python_graphs
    Download
    51885
  20. Bieber Babble@BieberBabble
    @BieberBabble
    [1] Learning to execute [2] Runtime error prediction [3] Global relational models of source code [4] Graph Finite-State Automaton [5] WIP Program Graphs
    115
  21. Bieber Babble@BieberBabble
    @BieberBabble
    For further research investigating how program structure can push #ML4Code forward, hopeful that python_graphs continues to prove useful!
    004
  27. Justin ("Goju") Gottschlich@j_gottschlich
    Hi @Stanford students - I'll be teaching CS329M, a new grad/adv undergrad course on machine programming, starting in Sept. 2022. Please register if you have interest in #AI/#ML for code (#ml4code)! More details here: Please RT!
    012
  29. Show more posts