📣 Today, we’re launching Provenance and Attribution, a new feature that reduces the risk of IP liability when using third-party models like Claude 3.5 Sonnet and GPT-4o. Get the details 👇
🔗
#AIforCode #SoftwareEngineers #DevTeams #LLMs
003323
Posts with #AIForCode
#AIForCode
Public posts on X that use #AIForCode.
Related hashtags
José A. Alonso@Jose_A_AlonsoIs AI making coders obsolete? (Are there problems with having AI tools take over coding from humans?). ~ Jennifer Goforth Gregory. #AIforCode007443
Diffblue@diffbluehq📰 We're delighted to share our latest news that Diffblue has been awarded an Innovate UK grant as part of the ITEA project Generative AI for the Software Development Life Cycle (GENIUS). 🔗Find out more: #Diffblue #news #AIforCode #AI #AIUK #UKBusiness
022134
Chaoyun Zhang@vyokkyProud to share our new work — SWE-bench-Live is now live! A live-updating benchmark for real-world bug fixing, where even top agents like Claude 3.7 Sonnet + OpenHands stumble. Try it out & follow us 👉 #LLM #SWEbenchLive #AIforCodeBowen Li @BowenLi2121🤔 Have we really made great progress on software engineering tasks? 🚀 Introducing SWE-bench-Live, a live-updatable benchmark for real-world bug fixing. 😺 Even the best combo, OpenHands + Claude 3.7 Sonnet, sees a major performance drop! 👉 🧵 1/4103239
- Updating a code library with AI and facing limitations with your LLM's knowledge? As ever, context matters. Here are some tips for making the process easier: 🔗 #AIforCode #Coding #developers #CodingChallenge
004453 - Diffblue@diffbluehqOne of the questions we get asked most frequently about Diffblue Cover is..."Why would I need Diffblue Cover for unit testing, if I have Copilot?" Find out the answers in the AI for unit testing showdown 🔗 #AI #AIforCode #UnitTesting #developertools #aicodingtools #AIcoding #CleanCode
0151.8K 
LocalAI@LocalAI_API🎉 New model alert! Check out "all-hands_openhands-lm-1.5b-v0.1" in LocalAI gallery! 🤖 Install it with `local-ai run all-hands_openhands-lm-1.5b-v0.1` and explore its potential for software engineering tasks! 💻🔥 #LocalAI #OpenHandsLM #AIforCode808194
Aditya Kanade@adityakanade0Can we assist code LMs reduce hallucinations in unfamiliar repos using static analysis? We present monitor-guided decoding (MGD) to evaluate this and find that it improves compilation rates and match with ground truth across the board. Arxiv: #AIforCode
AK @_akhaliqGuiding Language Models of Code with Global Context using Monitors paper page: Language models of code (LMs) work well when the surrounding code in the vicinity of generation provides sufficient context. This is not true when it becomes necessary to use types or functionality defined in another module or library, especially those not seen during training. LMs suffer from limited awareness of such global context and end up hallucinating, e.g., using types defined in other files incorrectly. Recent work tries to overcome this issue by retrieving global information to augment the local context. However, this bloats the prompt or requires architecture modifications and additional training. Integrated development environments (IDEs) assist developers by bringing the global context at their fingertips using static analysis. We extend this assistance, enjoyed by developers, to the LMs. We propose a notion of monitors that use static analysis in the background to guide the decoding. Unlike a priori retrieval, static analysis is invoked iteratively during the entire decoding process, providing the most relevant suggestions on demand. We demonstrate the usefulness of our proposal by monitoring for type-consistent use of identifiers whenever an LM generates code for object dereference. To evaluate our approach, we curate PragmaticCode, a dataset of open-source projects with their development environments. On models of varying parameter scale, we show that monitor-guided decoding consistently improves the ability of an LM to not only generate identifiers that match the ground truth but also improves compilation rates and agreement with ground truth. We find that LMs with fewer parameters, when guided with our monitor, can outperform larger LMs. With monitor-guided decoding, SantaCoder-1.1B achieves better compilation rate and next-identifier match than the much larger text-davinci-003 model.18294.9K
Guiding Language Models of Code with Global Context using Monitors paper page: Language models of code (LMs) work well when the surrounding code in the vicinity of generation provides sufficient context. This is not true when it becomes necessary to use types or functionality defined in another module or library, especially those not seen during training. LMs suffer from limited awareness of such global context and end up hallucinating, e.g., using types defined in other files incorrectly. Recent work tries to overcome this issue by retrieving global information to augment the local context. However, this bloats the prompt or requires architecture modifications and additional training. Integrated development environments (IDEs) assist developers by bringing the global context at their fingertips using static analysis. We extend this assistance, enjoyed by developers, to the LMs. We propose a notion of monitors that use static analysis in the background to guide the decoding. Unlike a priori retrieval, static analysis is invoked iteratively during the entire decoding process, providing the most relevant suggestions on demand. We demonstrate the usefulness of our proposal by monitoring for type-consistent use of identifiers whenever an LM generates code for object dereference. To evaluate our approach, we curate PragmaticCode, a dataset of open-source projects with their development environments. On models of varying parameter scale, we show that monitor-guided decoding consistently improves the ability of an LM to not only generate identifiers that match the ground truth but also improves compilation rates and agreement with ground truth. We find that LMs with fewer parameters, when guided with our monitor, can outperform larger LMs. With monitor-guided decoding, SantaCoder-1.1B achieves better compilation rate and next-identifier match than the much larger text-davinci-003 model.
32510144.3K- Diffblue@diffbluehqYou've been asking"What's the ROI of using AI for Code?". We've got you covered with the answer! Head over to our new, sleek homepage @ and we'll show you the math. #AIforCode #techstack #artificialintelligence
001 - José A. Alonso@Jose_A_AlonsoVERINA: Benchmarking verifiable code generation. ~ Zhe Ye, Zhengxu Yan, Jingxuan He, Timothe Kasriel, Kaiyu Yang, Dawn Song. #AIforCode #ITP #LeanProver016424
🧵 GROK is about to change the AI game this weekend: 5 groundbreaking features dropping now. Voice. Code. Vision. Multi-agents. And Tesla? Let’s dive in 👇
19206.9K- @Vicojims2️⃣ GROK 4 HEAVY 🧠 Think of it as multiple AIs debating before giving you the best answer. Multi-agent reasoning. Smarter responses. The first of its kind for public use. Available to SuperGrok Heavy subscribers. #Grok4 #AIRevolution
278795 - @Vicojims3️⃣ GROK 4 CODE (beta) 💻 Built just for devs: •256K token context •Smart debugging •Refactoring •Code suggestions An AI pair programmer on steroids. Now in private beta testing. #DevTool #AIForCode
178329 - Diffblue@diffbluehqWhen it takes so long to fit everyone on @diffblueHQ’s team update photo, even the office pooch has had enough. Find out more how we’re changing the way code is developed: #AIforCode #OfficeDog
013 
Daniel Kroening@DiffKroeningI'm often asked what does "AI for Code" mean, so I thought I'd write a quick post to explain the reasoning! Give it a read and let me know your thoughts! #AIforCode #ArtificialIntelligence
002- Diffblue@diffbluehqTake a sneak peek at our labs and see what alpha & beta products we're cooking up🧪 ➡️ #AIforCode #code
001 - Daniel Kroening@DiffKroeningThis is why, at @diffbluehq, we invest in hiring talented computer scientists and developers who may not have ever worked in AI before and training them to work with our #AI engine. #AIforCode #SoftwareDevelopment #startup #JavaVerdict @VerdictUKLabour markets all over the world can’t keep up with the demand for developers, mathematicians and scientists needed for AI. @diffbluehq035
Verdict@VerdictUKLabour markets all over the world can’t keep up with the demand for developers, mathematicians and scientists needed for AI. @diffbluehq024- Diffblue@diffbluehqWe've recently been told we're "the easy button" for testing. Check out to learn more. #aiforcode #developers #testing #Java #OneClick
101 - Diffblue@diffbluehqWhen you generate your #java tests on share the results with colleagues, friends, communities or really anyone you want to share results with. Just hit 'Share' after generating your test - it'll look something like this: #AIforCode013
- Diffblue@diffbluehqGetting frustrated with your #java #unittesting? Sure, you could @Google it - it'll just tell you to go to (it automates your unit tests for you). Even @googledevs would love it (probably). Go on, try it. You'll love it. #AIforCode #automatedtesting #AI
011 - Show more posts
