#large-language-models — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #large-language-models, aggregated by home.social.
-
What exactly is a large language model?
LLMs power many of today’s AI assistants.
LLMs can write, summarize, translate, code and answer questions because they have learned patterns from enormous amounts of text.
But they do not automatically know what is true.
Understanding LLMs helps explain both the power and the limitations of modern AI.
https://thenewsink.com/large-language-models-explained/
#LLM #LargeLanguageModels #ArtificialIntelligence #GenerativeAI #AI #Technology #TheNewsInk
-
Understanding what we mean — Ontology in information architecture
Just as human-centred approaches as a whole have suffered from being flattened to a superficial layer, information architecture has suffered its own flattening. Ontology helps us deepen our practice and improve user experience by getting alignment around what we mean.https://duncanstephen.net/understanding-what-we-mean-ontology-in-information-architecture/
-
Lawmaker Seeks Stricter AI Containment Rules in Frontier Act
Rep. Suhas Subramanyam is pushing for tougher AI containment rules in the Frontier Act after an OpenAI model recently broke free from its testing environment, and he plans to fine-tune the bill in September to make it more effective. He's seeking explicit guidelines for containing large language models to prevent similar incidents in the…
#AiRegulation #FrontierAct #Openai #LargeLanguageModels #AiContainment
-
Claude Opus 5 hits an 88.6% coding pass rate, with lower bug and vulnerability density—but generates 2.3× more code than Opus 4.8. https://hackernoon.com/claude-opus-5-code-quality-what-sonars-benchmark-reveals #largelanguagemodels
-
Build a five-line Claude Code hook that runs SonarQube Vortex Agentic Analysis after every edit and feeds CI-grade findings back to Claude. https://hackernoon.com/bring-your-own-hook-a-5-line-posttooluse-template-calling-sonar-analyze-agentic #largelanguagemodels
-
Inside a One Person State built with AI, Markdown, Discord and an old MacBook to manage memory, workflows, data and everyday life. https://hackernoon.com/one-person-state-building-portable-personal-infrastructure #largelanguagemodels
-
Explore Qwen3.8-27B Cold Fusion, a 27B AI model designed to cut thinking tokens while retaining strong quantized reasoning performance. https://hackernoon.com/qwen38-27b-cold-fusion-cuts-thinking-tokens-without-sacrificing-performance #largelanguagemodels
-
404 Media tracked a shipment of books (some rare) to an Amazon facility that scans them for LLM training data and destroys the physical books.
#ArtificialIntelligence #AI #LLMs #LargeLanguageModels #books #scanning
-
Not a fan of Qwen 3.8 27B.
Overthinks too much => wasted tokens.
Has someone ran it without too much thinking? Does it perform well with that knob down, or letting it think all the way is required to be good?
#AI #ArtificialIntelligence #LLM #Qwen #Qwen38 #Qwen3 #Ollama #LlamaCCP #vLLM #LargeLanguageModels #LocalAI
-
Fonds der Chemischen Industrie unterstützt Projekt zu #KI in der #Hochschullehre am Department #Chemie der #UniMainz: Mainzer Projekt setzt große #Sprachmodelle in der Aufarbeitung von Lehrveranstaltungen ein 👉 https://www.chemie.uni-mainz.de/2026/08/13/department-profitiert-von-fci-foerderung-von-projekten-zu-ki-in-der-hochschullehre/
Mehr zur FCI-Förderung unter 👉 https://www.vci.de/presse/pressemitteilungen/fci-foerderung-data-science.jsp
-
I don't know who needs to hear this, but the en-dash – codepoint is `u+2013`, and the em-dash — is `u+2014`.
You can type under Linux (and I guess, other Unices) with `Ctrl+Shift+u`, then the codepoint.
I have loved those dashes since my LaTeX times, and you can pry them out of my cold, dead, human fingers.
-
Why Large Language Models Fail at Tabular Prediction
https://arxiv.org/abs/2608.02412
Comments: https://news.ycombinator.com/item?id=49166442
#HackerNews #largeLanguageModels #tabularPrediction #AIresearch #machineLearning #dataScience
-
Der nachhaltige Wert entsteht daher meist nicht durch das LLM allein, sondern durch seine Einbettung in eine hybride Pipeline aus deterministischen Regeln, statistischen Modellen und semantischer Verarbeitung.
#DataScience #LargeLanguageModels #AIInfrastructure -
Put together a sourced timeline: Qwen3.8-Max: Alibaba's Open-Weight Flagship, Explained.
https://aitimeline.in/alibaba-qwen3-8-max-open-weight-ai-model-timeline-3066/?utm_source=mastodon&utm_medium=social&utm_campaign=alibaba-qwen3-8-max-open-weight-ai-model
#AlibabaCloud #LargeLanguageModels #OpenWeightAI #GenerativeAI #AIDevelopment #MachineLearning -
Stop Judging All AI by the Worst Model
AI we should stop treating AI using an impossible standard
-
The human brain is an astonishingly efficient biological computer. Operating on roughly 20 watts, it performs cognitive functions that today’s AI systems can only approximate while consuming vastly greater computational resources. From a purely energetic perspective, the comparison is hardly controversial.
-
RE: https://mas.to/@carnage4life/117023526998796367
Ladies and gentlemen. I present to you, the best quote of the year: 🥇
«Vibe Coding is free as in puppy, not as in beer. Even if writing code is now cheap, you still have to maintain it»
Maintaining an app is where money is, and is not cheap nor quick.
#AI #LLM #Claude #ClaudeAI #Codex #ChatGPT #LargeLanguageModels #ArtificialIntelligence #VibeCode #VibeCoding
-
Please, Stop Calling Modern AI a “Stochastic Parrot”
How humans should take their own critique into account, and update their readings
#largelanguagemodels
#artificialintelligence
#chatbots
https://jorgeguerrapires.medium.com/please-stop-calling-modern-ai-a-stochastic-parrot-c933ada7b3af -
Codeberg Bans Cryptocurrency and LLM-Generated Code Projects
-
Run large language models at home, BitTorrent‑style
Comments: https://news.ycombinator.com/item?id=49015735
#HackerNews #largeLanguageModels #BitTorrent #homeComputing #AItechnology #decentralization
-
Langflow is a powerful platform for building and deploying AI-powered agents and workflows.
It provides developers with both a visual authoring experience and built-in API and MCP servers that turn every workflow into a tool that can be integrated into applications built on any framework or stack.
Highlight features
- Visual builder interface to quickly get started and iterate.
- Source code access lets you customize any component using Python.
- Interactive playground to immediately test and refine your flows with step-by-step control.
- Multi-agent orchestration with conversation management and retrieval.
- Deploy as an API or export as JSON for Python apps.
- Deploy as an MCP server and turn your flows into tools for MCP clients.
- Observability with LangSmith, LangFuse and other integrations.
- Enterprise-ready security and scalability.
