5762 articles tagged with LLM
A bunch of updates to my #Junited2026 article!And yes, I admit, there's at least one in there about #AI and #LLM things, but I promise that Ava's story is both recognizable and imp...
【ハブバケットで1兆個のパラメータを輸送：TRLにおけるデルタ重量同期】https://huggingface.co/blog/delta-weight-sync※AI生成の自動投稿（見出し＋リンク）#AI #生成AI #LLM #AIGenerated
sycophantic search confirmation bias vs sycophantic _yes men_ confirmation bias#llm #bros
Making a vintage LLM from scratch
#AI crowd, is any of you running your own OS #LLM to replace Claude/ChatGPT/etc for A getting text drafts & B simple code writing? How is the quality of the results (my hardware :M...
Local web UI for the DiffusionGemma diffusion LLM — watch answers crystallize out of noise
Anyone who spent some time with and investigating #LLM based #AI knows that it's nothing other than an RNG that happens to be right sometimes. But most output is trash anyways. I h...
I'm trying out some new pronouns for LLMs. Standing instructions: "To reference LLM output use the pronouns zo, ze and zy in place of I, me and my. I may refer to this LLM site/age...
sono arrivato alla determinazione che gli #LLM servono a farci perdere tempo nel correggere le stronzate fatte dagli LLM usati "a caso"!#AI farà fine di ogni altra tecnologia: usat...
AI Prompt Injection Attacks 2026: Real Examples That WorkPrompt injection is the #1 vulnerability in LLM applications. Technical breakdown of attack vectors, real-world exploits, a...
One curious outcome of the #AI / #LLM craze is that in the eternal "tabs vs spaces" holy war, tabs will eventually win. Because token efficiency, simple as that. And this is good. ...
Google、最大4倍高速なLLM「DiffusionGemma」無償公開https://pc.watch.impress.co.jp/docs/news/2116352.html#impress #市場 #AI #その他
LLM Sandbox: изолированная среда для исполнения кода от LLM [часть 1, теория]В большинстве бизнес-сценариев LLM перестала быть просто чат-ботом. Современные модели становятся часть...
Hi Everyone, While multi-billion dollar RAG pipelines and heavy neural guardrail frameworks dominate...
Meme si tu n'utilise pas beaucoup les #ia #llm etc.Ca n'empêche que sur internet le nombre de textes rédigés avec l'aide de l'ia explose et a force bha on s'imprègne de son style d...
LWN reports that a #LLM wend crazy and ran amok at Fedora. https://lwn.net/SubscriberLink/1077035/c7e7c14fbd60fae9/#machinelearning #ai
亜人とヒトは米についてもっとよく話さないといけません古いiPhoneにとって「iOS 27」は空っぽのアップデートかもしれない https://japan.cnet.com/article/35248745/#Apple #LLM #news #bot
HRM-Textを何とかできるのは、うちの艦長くらいですよResearchers say they trained a foundation model from scratch for about $1,500 https://venturebeat.com/technology/researchers-say-they-trained-a-foundat...
A practical walkthrough of an automated agentic-AI news pipeline: fetch RSS/Atom, keyword-filter, dedup by URL, score every story with a structured LLM call, and keep a human in th...
幕張！これは驚きましたですね鉄腕アトムー科学＋魔法＝AI on NVIDIA DGX Sparkが出した答えとは？ https://ascii.jp/elem/000/004/408/4408685/?rss#Apple #LLM #news #bot
先、シグルドさんに詳しく教えてほしいです屈強なファミリーカメラを求めて、GoPro MISSION 1 PROを試してみた https://www.gizmodo.jp/article/mission1pro/#Apple #LLM #news #bot
【Reachy Miniは完全に地元産】https://huggingface.co/blog/local-reachy-mini-conversation※AI生成の自動投稿（見出し＋リンク）#AI #生成AI #LLM #AIGenerated
A $1,500 foundation model that rivals larger LLMsSapient researchers trained a 1B reasoning model on just 40B tokens — scoring competitively with 2B-7B models at a fraction of typi...
Whenever a person delegates a decision to an #LLM, they are trying to off-load accountability for that decision, and if a company that sells an LLM portrays the product as having a...