#machine-learning — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #machine-learning, aggregated by home.social.
-
Season 1 Lesson 38 Part 7 - Your First Steps in Python Python Inheritance Classes Isinstance #machinelearning #codingtutorial #dataanalysis #softwarengineer #dataengineer #softwaredeveloper #gcp #azure #jupyternotebook #pythoncode #pythonprogramming #datascience #vibecoding #python #aws #learncoding
-
Claude Academy est en ligne : cours et tutoriels gratuits d'Anthropic pour apprendre à utiliser Claude, du niveau débutant (comprendre l'IA) jusqu'à l'intégration en équipe via l'API, Claude Code ou MCP. ⬇️
https://academy.claude.com/📬 Ma veille dev de la semaine → https://l.camilleroux.com/veille-aFv
-
Simile AI raised a $2B Series B, backed by GreenOaks and Index Ventures, running tens of millions of simulations for Fortune 100 clients like CVS with 85-99% accuracy vs human focus groups.
Source: Latent Space
https://www.latent.space/p/simile -
📄 Paper alert: EnvHarness: Awakening Static Worlds for Agent Learning — 235 upvotes on Hugging Face. It turns static worlds into interactive learning environments, a bridge between offline data and agent training.
-
In 2003, Paul Eilers published a paper titled “A Perfect Smoother” [1], re-implementing and extending early ideas published by Whittaker in the 1920s [2].
The method is based on a penalised least square approach, also called Tikhonov regularization (also called ridge regression!) and is a very general approach to deal with smoothing noisy data, overcoming some of the limitations of the Savitzky-Golay method.
Read more on my recent blog post 👇
https://nirpyresearch.com/whittaker-smoothing/
[1] Paul H. C. Eilers (2003). A Perfect Smoother, Anal. Chem. 75 (14): 3631–3636.
[2] E. T. Whittaker (1922). On a New Method of Graduation, Proceedings of the Edinburgh Mathematical Society, 41: 63 – 75.#spectroscopy #smoothing #DataProcessing #ImageProcessing #Python #MachineLearning #Regression
-
AI tools fade into the background when they become routine. I built a RAG system that saved me 3.5 hours per QA cycle. Then I stopped noticing it.
To fix it, I started treating AI like a junior teammate. New tasks daily, outside my usual workflow. Found 3 edge cases in my pipeline I’d overlooked.
Full write-up: https://www.adilaidev.com/blog/ai-blindness-is-real-heres-how-to-fix-it
-
Season 1 Lesson 38 Part 6 - Your First Steps in Python Python Inheritance Dog Cat Test #datascience #dataengineer #softwaredeveloper #pythoncode #machinelearning #codingtutorial #dataanalysis #softwarengineer #pythonprogramming #vibecoding #jupyternotebook #gcp #azure #python #aws #learncoding
-
Reddit is hiring Machine Learning Manager, Feed Ecosystems
🔧 #machinelearning
🌎 Remote; United States
⏰ Full-time
🏢 RedditJob details https://jobsfordevelopers.com/jobs/machine-learning-manager-feed-ecosystems-at-redditinc-com-jul-29-2026-cc1610?utm_source=mastodon.world&utm_medium=social&utm_campaign=posting
#jobalert #jobsearch #hiring -
Put together a sourced timeline: AI Evolution Timeline: ChatGPT to Autonomous Coding Agents.
https://aitimeline.in/ai-evolution-timeline-chatgpt-autonomous-coding-agents-5420/?utm_source=mastodon&utm_medium=social&utm_campaign=ai-evolution-timeline-chatgpt
#ArtificialIntelligence #ChatGPT #AutonomousAgents #GPT4 #MachineLearning #Technology -
How we made a text-to-speech model respond in sub-50 ms
https://nari-labs.com/blog/qwen3-tts-speed-cost-frontier/
Comments: https://news.ycombinator.com/item?id=49389952
#HackerNews #texttospeech #machinelearning #AIperformance #speechsynthesis
-
FYI: Explaining GPTBot: GPTBot is OpenAI's training crawler for foundation models. The robots.txt token, user agent versions, published IP ranges, blocking rates and the open disputes. https://ppc.land/gptbot/ #GPTBot #OpenAI #AI #MachineLearning #DataScience
-
SOP-Bench is a new benchmark for evaluating AI agents on real business procedures. It pairs genuine enterprise SOPs with functioning tools and ground-truth answers.
Source: Amazon Science
https://www.amazon.science/blog/sop-bench-a-new-benchmark-for-evaluating-ai-agents-on-real-business-procedures -
Insane Prompts The Ultimate Prompt Blueprints To Become a Business Leader
With 100 carefully crafted, high-impact prompts, you’ll start seeing AI deliver results that actually move the needle. We’re talking smarter content, better marketing angles, and even faster workflows. These aren’t your average prompts they’re designed to make AI think deeper, work harder, and help you achieve more, whether you’re creating content, building funnels, or just trying to get ahead. These aren’t random or recycled they’re purpose-built to help you grow in key areas […] -
A two-week randomised trial with first-year university students found that only those assigned to text a randomly paired peer reported less loneliness at the end. Students assigned to a supportive chatbot called Sam did not differ from the journal group in loneliness reduction, though the AI reduced negative mood.
Source: Silicon Canals
https://siliconcanals.com/t-ai-friend-human-peer-loneliness-randomised-study/ -
SkillOpt (Microsoft) : un optimizer de skills en langage naturel pour agents LLM. Le skill s'améliore via des rollouts scorés, sans toucher aux poids du modèle. Le fichier best_skill.md est portable d'un modèle à l'autre. Open source, MIT. ⬇️
https://github.com/microsoft/SkillOpt📬 Ma veille dev de la semaine → https://l.camilleroux.com/veille-zJr
-
Explainable 'AI' using Gradient Boosted randomized networks Pt2 (the Lasso)
https://thierrymoudiki.github.io/blog/2020/07/31/python/r/lsboost/explainableml/mlsauce/xai-boosting-2