#LM

José-María Súnicojmsunico@fosstodon.org
2026-02-12

sn-news: #open #sw #dev #ai #ml #lm #hmc Vibe Coding Kills Open Source arxiv.org/pdf/2601.15494

Gaza_campkids🇵🇸rouh
2026-01-20

G'morning 💕 , guys plz wish me luck 🥹
Today I have an interview for a scholarship to a data science training program with Heal Palestine, and I'm currently training for it. If anyone has any advice or tips, please share them with me.
@aral @fabio my legend in tech , do you have any advice or info I should know before the interview?

2026-01-14

Does anyone know how to get an R-squared value for a fitted model using #lavaan in #Rstats? Specifically, I'm looking for R-squared for a regression model (i.e., one DV). I'm finding very little to work with online. I have a coauthor insisting that R2 is critical for his publication venue.

#question #lm #model #sem #statistics

José-María Súnicojmsunico@fosstodon.org
2026-01-13
2025-12-27

[…] Contrary to how it may seem when we observe its output, an LM is a system for haphazardly stitching together sequences of linguistic forms it has observed in its vast training data, according to probabilistic information about how they combine, but without any reference to meaning: a stochastic parrot.

“On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?” by Bender, Gebru, McMillan-Major, and Mitchell (2021)

s10251.pcdn.co/pdf/2021-bender

@ngaylinn #ai #degenerativeai #lm

miciomaomiciomao
2025-12-18
José-María Súnicojmsunico@fosstodon.org
2025-12-14

sn-news: #open #ai #ml #lm #hmc Linux Foundation Announces the Formation of the Agentic AI Foundation (AAIF), Anchored by New Project Contributions Including Model Context Protocol (MCP), goose and AGENTS.md linuxfoundation.org/press/linu

Miikka VikkulaMiikka_Vikkula
2025-12-05

So great 2 days at the Society of 2025 in Toulouse on lymphatic malformations, lymphoedema and lipedema. Illustrative presentations, pertinent questions and collegial discussions. The field is moving along faster than ever. Humbling to be part of the faculty.

2025-12-05

Một người dùng Reddit đang đặt câu hỏi về khả năng sử dụng thuật toán fast-wavenet (vốn dùng tạo âm thanh) để xây dựng mô hình suy luận cho Mô hình Ngôn ngữ (LM). Liệu ý tưởng này khả thi và hiệu quả ra sao?

#AI #Wavenet #LM #MachineLearning #ArtificialIntelligence #MôHìnhNgônNgữ #HọcMáy

reddit.com/r/LocalLLaMA/commen

Le Néandertal se sent las, lasHydrePrever@mathstodon.xyz
2025-12-04

Currently teaching lm/glm.

Does anyone know a real life (publicly available) dataset for which the fitted value vs residuals plot would look like this? (quantitative response y; as many predictors as you want, here it was generated with a single one but if there are several, it's ok)

🔁 welcome

#stats #lm #dataset #rstats #teaching #academicChatter

a diagnostic plot in which there's a very apparent structure of the residuals, which seem to have a standard deviation roughly proportional to the predicted value
2025-11-25

Langchain đang tích hợp LM Studio như một trình điều khiển cục bộ mới, tương tự cách Ollama được sử dụng với `from langchain_ollama.llms import OllamaLLM`. Người dùng LM Studio có thể cần sử dụng phương pháp OpenAI hoặc tạo endpoint cho mô hình cục bộ để kết nối với Langchain. Cách tiếp cận này giúp phát triển các agent AI cục bộ hiệu quả. #AI #Langchain #LM Studio #LocalLLM #AI VietNam #CongNghe #MachineLearning

reddit.com/r/LocalLLaMA/commen

2025-11-01

ChatGPT đang mất dần tính hữu ích từ 2025, rào cản nghiêm ngặt làm giảm tính năng hỗ trợ pháp lý, y tế và thảo luận cá nhân. Mô hình mới 5 kém hiệu quả, đánh dấu giai đoạn "thuần hóa" AI. Tự do LLM đã chấm dứt. #ChatGPT #AISecurity #AI #LM #EnsGPT

reddit.com/r/singularity/comme

José-María Súnicojmsunico@fosstodon.org
2025-10-20

sn-news: #open #ch #ai #ml #lm #hmc Apertus: a fully open, transparent, multilingual language model actu.epfl.ch/news/apertus-a-fu

2025-10-14

NVIDIA DGX Spark: great hardware, early days for the ecosystem - NVIDIA sent me a preview unit of their new DGX Spark desktop "AI supercompute... - simonwillison.net/2025/Oct/14/ #generative-ai #coding-agents #claude-code #disclosures #local-llms #tailscale #llama-cpp #lm-studio #hardware #nvidia #ollama #llms #ai

Stefanie Janine Stöltingsjstoelting@digitalcourage.social
2025-09-30

@heiseonline Geniale Idee, KI auf Krankenakten loslassen.
Die Dokumentationen werden dann genauso gut, wie die anderen Lügen der Lie Machines (#LM).
Ganz davon abgesehen, dass die Daten schon mit der Erfassung direkt ausgeleitet werden können.

Client Info

Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst