#parameter — Public Fediverse posts

Квадриллион параметров (1Q) для LLM. Когда нам ждать такую модель?

Согласно закону заголовков Беттериджа , который утверждает: «Если заголовок заканчивается вопросительным знаком, на него можно ответить „нет“», ответом на вопрос в заголовке будет „никогда“. И я попытаюсь вас в этом убедить.

https://habr.com/ru/articles/1065320/

#ии #ииагенты #ai #chatgpt #claude #kimi #parameter #dataset #billion

Critical Copilot vulnerability allowed hackers to steal 2FA code from users

SearchLeak exploit shows why the industry’s approach to LLM security fails over and over.

Archive: ia: https://s.faithcollapsing.com/wuxba

#ai #biz-&-it #copilot #llms #parameter-to-prompt-injection #security
https://arstechnica.com/security/2026/06/critical-copilot-vulnerability-allowed-hackers-to-seal-2fa-code-from-users/

Critical Copilot vulnerability allowed hackers to steal 2FA code from users

SearchLeak exploit shows why the industry’s approach to LLM security fails over and over.

Archive: ia: https://s.faithcollapsing.com/wuxba

#ai #biz-&-it #copilot #llms #parameter-to-prompt-injection #security
https://arstechnica.com/security/2026/06/critical-copilot-vulnerability-allowed-hackers-to-seal-2fa-code-from-users/

Critical Copilot vulnerability allowed hackers to seal 2FA code from users

SearchLeak exploit shows why the industry’s approach to LLM security fails over and over.

#ai #biz-&-it #copilot #llms #parameter-to-prompt-injection #security
https://arstechnica.com/security/2026/06/critical-copilot-vulnerability-allowed-hackers-to-seal-2fa-code-from-users/

Critical Copilot vulnerability allowed hackers to seal 2FA code from users

SearchLeak exploit shows why the industry’s approach to LLM security fails over and over.

#ai #biz-&-it #copilot #llms #parameter-to-prompt-injection #security
https://arstechnica.com/security/2026/06/critical-copilot-vulnerability-allowed-hackers-to-seal-2fa-code-from-users/

#fedihelp

in #youtube urls the #parameter si= is known.

but now i have seen is=

what is that? where does it come from?

#fedihelp

in #youtube urls the #parameter si= is known.

but now i have seen is=

what is that? where does it come from?

Firefox irritations

The sponsored links enabled on the homescreen by default irritated me some odd days ago. I saw them when I checked a installation on a new Android installation, where I had disabled them, and was also very aware that I had only  patched the Android versions accross all mobile, and X86 elf versions across desktop devices.
It was a deliberate design choice to reset certain settings, by the Firefox programmers.
They had also reset ffox DNS feature, parameters, which make no sense, since my ISP DNS is the closest and all my traffic is encrypted anyway

All those devices had that particular parameter reset {including some local LLM features}

I'm willing to bet that something similar also happened on BSD machines

I had to put the disclaimer for the language I used in that particular toot ;)

It is unfortunate that the forks also seem to get similar problems, just due to the sheer volume of changes that a massive project like Firefox brings with it.

I've said this decade ago and I repeat it

One person cannot write a massive undertaking which is called a browser but in reality is a whole operating system which should be in a sandbox

@rl_dane

#Firefox #parameter #reset #after #patch #programming #Android #LLM #AI #slop #Linux #DNS #BSD

Firefox irritations

The sponsored links enabled on the homescreen by default irritated me some odd days ago. I saw them when I checked a installation on a new Android installation, where I had disabled them, and was also very aware that I had only  patched the Android versions accross all mobile, and X86 elf versions across desktop devices.
It was a deliberate design choice to reset certain settings, by the Firefox programmers.
They had also reset ffox DNS feature, parameters, which make no sense, since my ISP DNS is the closest and all my traffic is encrypted anyway

All those devices had that particular parameter reset {including some local LLM features}

I'm willing to bet that something similar also happened on BSD machines

I had to put the disclaimer for the language I used in that particular toot ;)

It is unfortunate that the forks also seem to get similar problems, just due to the sheer volume of changes that a massive project like Firefox brings with it.

I've said this decade ago and I repeat it

One person cannot write a massive undertaking which is called a browser but in reality is a whole operating system which should be in a sandbox

@rl_dane

#Firefox #parameter #reset #after #patch #programming #Android #LLM #AI #slop #Linux #DNS #BSD

Sidenote: if the #automated knowledge (e.g., a #parameter to reduce the search space) is very similar to the human input, and if the researcher holds a #positivist view on things -- they are epistemically welcome to use this additional parameter for their research/ in their pipeline.
I would argue that they shouldnt boast with having used/integrated #qualitativeData though, it is a parameter learned from (in most cases) textual data.

Sidenote: if the #automated knowledge (e.g., a #parameter to reduce the search space) is very similar to the human input, and if the researcher holds a #positivist view on things -- they are epistemically welcome to use this additional parameter for their research/ in their pipeline.
I would argue that they shouldnt boast with having used/integrated #qualitativeData though, it is a parameter learned from (in most cases) textual data.

Run a 1T parameter model on a 32gb Mac by streaming tensors from NVMe

https://github.com/t8/hypura

#HackerNews #Run #a #1T #parameter #model #on #a #32gb #Mac #by #streaming #tensors #from #NVMe #https://github.com/t8/hypura #MachineLearning #Tensors #NVMe #Mac #Optimization

Run a 1T parameter model on a 32gb Mac by streaming tensors from NVMe

https://github.com/t8/hypura

#HackerNews #Run #a #1T #parameter #model #on #a #32gb #Mac #by #streaming #tensors #from #NVMe #https://github.com/t8/hypura #MachineLearning #Tensors #NVMe #Mac #Optimization

Code Smells sind Hinweise darauf, dass im Code etwas nicht sauber ist - etwas schlecht riecht. Es geht nicht um Syntaxfehler oder Bugs, sondern um Strukturen, die dich langfristig ausbremsen. Der Code funktioniert vielleicht heute, aber er wird schwerer zu verstehen, zu testen und zu erweitern. Gera

https://magicmarcy.de/code-smells-was-riecht-denn-hier-so-streng

#CodeSmells #Methoden #Logik #Parameter #MagicStrings #MagicNumbers #Programming #Awareness

LLMs contain a LOT of parameters. But what’s a parameter? – MIT Technology Review

Artificial intelligence

LLMs contain a LOT of parameters. But what’s a parameter?

They’re the mysterious numbers that make your favorite AI models tick. What are they and what do they do?

By Will Douglas Heavenarchive page

January 7, 2026

Photo Illustration by Sarah Rogers/MITTR | Photos Getty

MIT Technology Review Explains: Let our writers untangle the complex, messy world of technology to help you understand what’s coming next. You can read more from the series here.

I am writing this because one of my editors woke up in the middle of the night and scribbled on a bedside notepad: “What is a parameter?” Unlike a lot of thoughts that hit at 4 a.m., it’s a really good question—one that goes right to the heart of how large language models work. And I’m not just saying that because he’s my boss. (Hi, Boss!)

A large language model’s parameters are often said to be the dials and levers that control how it behaves. Think of a planet-size pinball machine that sends its balls pinging from one end to the other via billions of paddles and bumpers set just so. Tweak those settings and the balls will behave in a different way.  

OpenAI’s GPT-3, released in 2020, had 175 billion parameters. Google DeepMind’s latest LLM, Gemini 3, may have at least a trillion—some think it’s probably more like 7 trillion—but the company isn’t saying. (With competition now fierce, AI firms no longer share information about how their models are built.)

But the basics of what parameters are and how they make LLMs do the remarkable things that they do are the same across different models. Ever wondered what makes an LLM really tick—what’s behind the colorful pinball-machine metaphors? Let’s dive in.  

What is a parameter?

Think back to middle school algebra, like 2a + b. Those letters are parameters: Assign them values and you get a result. In math or coding, parameters are used to set limits or determine output. The parameters inside LLMs work in a similar way, just on a mind-boggling scale. 

Editor’s Note: Read the rest of the story, at the below link.

Continue/Read Original Article Here: LLMs contain a LOT of parameters. But what’s a parameter? | MIT Technology Review

LLMs contain a LOT of parameters. But what’s a parameter? – MIT Technology Review

Artificial intelligence

LLMs contain a LOT of parameters. But what’s a parameter?

They’re the mysterious numbers that make your favorite AI models tick. What are they and what do they do?

By Will Douglas Heavenarchive page

January 7, 2026

Photo Illustration by Sarah Rogers/MITTR | Photos Getty

MIT Technology Review Explains: Let our writers untangle the complex, messy world of technology to help you understand what’s coming next. You can read more from the series here.

I am writing this because one of my editors woke up in the middle of the night and scribbled on a bedside notepad: “What is a parameter?” Unlike a lot of thoughts that hit at 4 a.m., it’s a really good question—one that goes right to the heart of how large language models work. And I’m not just saying that because he’s my boss. (Hi, Boss!)

A large language model’s parameters are often said to be the dials and levers that control how it behaves. Think of a planet-size pinball machine that sends its balls pinging from one end to the other via billions of paddles and bumpers set just so. Tweak those settings and the balls will behave in a different way.  

OpenAI’s GPT-3, released in 2020, had 175 billion parameters. Google DeepMind’s latest LLM, Gemini 3, may have at least a trillion—some think it’s probably more like 7 trillion—but the company isn’t saying. (With competition now fierce, AI firms no longer share information about how their models are built.)

But the basics of what parameters are and how they make LLMs do the remarkable things that they do are the same across different models. Ever wondered what makes an LLM really tick—what’s behind the colorful pinball-machine metaphors? Let’s dive in.  

What is a parameter?

Think back to middle school algebra, like 2a + b. Those letters are parameters: Assign them values and you get a result. In math or coding, parameters are used to set limits or determine output. The parameters inside LLMs work in a similar way, just on a mind-boggling scale. 

Editor’s Note: Read the rest of the story, at the below link.

Continue/Read Original Article Here: LLMs contain a LOT of parameters. But what’s a parameter? | MIT Technology Review

Everybody who wants to understand Generative AI should read articles like this.
https://www.technologyreview.com/2026/01/07/1130795/what-even-is-a-parameter
#AI #parameter

Everybody who wants to understand Generative AI should read articles like this.
https://www.technologyreview.com/2026/01/07/1130795/what-even-is-a-parameter
#AI #parameter

https://www.technologyreview.com/2025/05/20/1116327/ai-energy-usage-climate-footprint-big-tech

This is an actually good article about #AI #energy use, and it's affect on #carbon, and #ClimateChange.

Here are some key things I would add.

All of these #corporations like #OpenAI #Microsoft and #Google are relying on an #LLM being accessed from the #cloud (the #internet).

#NVIDIA is counting on every increasingly big models.

Do I think this is the future of #generativeai

No.

I think _that_ is a big #Bubble. I think every "Size Up" on an AI model gives you an extra 20% in quality.

So, running a LLAMA 8b is only 20% better than running a LLAMA 3b.

Right?

So what?

Well, 99% of the usecase people have don't require 1 Trillion #parameter models.

They require models that increasingly can be run locally.

What do I think is coming? It isn't 30gigawatt data centers, it's laptops running on less energy than current ones with a terrabyte of RAM.

No one wants the #internet that the #MegaCorporations have created. People want their own shit. They want to own their movies.

The #cloud is dead. The companies built on the cloud are dead.

The just don't know it yet.

@mittechreview

https://www.technologyreview.com/2025/05/20/1116327/ai-energy-usage-climate-footprint-big-tech

This is an actually good article about #AI #energy use, and it's affect on #carbon, and #ClimateChange.

Here are some key things I would add.

All of these #corporations like #OpenAI #Microsoft and #Google are relying on an #LLM being accessed from the #cloud (the #internet).

#NVIDIA is counting on every increasingly big models.

Do I think this is the future of #generativeai

No.

I think _that_ is a big #Bubble. I think every "Size Up" on an AI model gives you an extra 20% in quality.

So, running a LLAMA 8b is only 20% better than running a LLAMA 3b.

Right?

So what?

Well, 99% of the usecase people have don't require 1 Trillion #parameter models.

They require models that increasingly can be run locally.

What do I think is coming? It isn't 30gigawatt data centers, it's laptops running on less energy than current ones with a terrabyte of RAM.

No one wants the #internet that the #MegaCorporations have created. People want their own shit. They want to own their movies.

The #cloud is dead. The companies built on the cloud are dead.

The just don't know it yet.

@mittechreview

Just had an interesting (but a bit unsettling) conversation with #AI about what it thought its future abilities would be. It’s a long response but worth the read.

My question: What new #emergent #abilities will occur when #parameter #numbers can be increased by orders of #magnitude?

AI’s response: This is a fascinating and actively researched question in AI! As language models like GPT grow in parameter count by orders of magnitude (e.g., from billions to tens or hundreds of trillions),

1/8

Just had an interesting (but a bit unsettling) conversation with #AI about what it thought its future abilities would be. It’s a long response but worth the read.

My question: What new #emergent #abilities will occur when #parameter #numbers can be increased by orders of #magnitude?

AI’s response: This is a fascinating and actively researched question in AI! As language models like GPT grow in parameter count by orders of magnitude (e.g., from billions to tens or hundreds of trillions),

1/8

A cycloidal pendulum - one suspended from the cusp of an inverted cycloid - is isochronous, meaning its period is constant regardless of the amplitude of the swing. Please find the proof using energy methods: Lagrange's equations (in the images attached to the reply).

Background:
The standard pendulum period of \(2\pi\sqrt{L/g}\) or frequency \(\sqrt{g/L}\) holds only for small oscillations. The frequency becomes smaller as the amplitude grows. If you want to build a pendulum whose frequency is independent of the amplitude, you should hang it from the cusp of a cycloid of a certain size, as shown in the gif. As the string wraps partially around the cycloid, the effect decreases the length of the string in the air, increasing the frequency back up to a constant value.

In more detail:
A cycloid is the path taken by a point on the rim of a rolling wheel. The upside-down cycloid in the gif can be parameterized by \((x, y)=R(\theta-\sin\theta, -1+\cos\theta)\), where \(\theta=0\) corresponds to the cusp. Consider a pendulum of length \(L=4R\) hanging from the cusp, and let \(\alpha\) be the angle the string makes with the vertical, as shown (in the proof).

#Pendulum #Cycloid #Period #Frequency #SHM #TimePeriod #CycloidalPendulum #Lagrange #Cusp #Energy #KineticEnergy #PotentialEnergy #Lagrangian #Length #Math #Maths #Physics #Mechanics #ClassicalMechanics #Amplitude #CircularFrequency #Motion #Vibration #HarmonicMotion #Parameter #ParemeterizedEquation #GoverningEquations #Equation #Equations #DifferentialEquations #Calculus

#open #call for #generative #art

~~~~~

#Parameter Art Fair 2025 organized by #Ljudmila during #Adela Festival, #Ljubliana calls for participation.

Parameters explores the practices arising in the expanded field of generative arts and critically probes the material and cultural aspects of generative systems to counter the dominant techno-cultural currents and narratives surrounding generative art and technology.

#open #call for #generative #art

~~~~~

#Parameter Art Fair 2025 organized by #Ljudmila during #Adela Festival, #Ljubliana calls for participation.

Parameters explores the practices arising in the expanded field of generative arts and critically probes the material and cultural aspects of generative systems to counter the dominant techno-cultural currents and narratives surrounding generative art and technology.

Заметки по ROS: параметры и Parameter server

Это третья из серии статей о робототехнической платформе ROS . Первую статью можно найти здесь . В этой статье я расскажу о параметрах и Parameter server в ROS. Кому интересно прошу под кат.

https://habr.com/ru/articles/871944/

#ros #parameter #configuration

So I've noticed my #LLM #HomeAssistant sometimes signing its #notifications. But it doesn't have a name, so it will either #hallucinate one or use my name or my wife's name. I could try to get it to just not do that, but it is a #7B #parameter model, so it often will try to comply but get hung up on the "structure" of what it thinks it is doing. So rather than beating it down, I was thinking I should just name it.

Any #fun names for an #automated home assistant? :wizard:

So I've noticed my #LLM #HomeAssistant sometimes signing its #notifications. But it doesn't have a name, so it will either #hallucinate one or use my name or my wife's name. I could try to get it to just not do that, but it is a #7B #parameter model, so it often will try to comply but get hung up on the "structure" of what it thinks it is doing. So rather than beating it down, I was thinking I should just name it.

Any #fun names for an #automated home assistant? :wizard:

Last plot before taking the train to Ljubljana #CreativeCoding #penplotter #adela #parameter

Last plot before taking the train to Ljubljana #CreativeCoding #penplotter #adela #parameter

Property Parameter Argument の違い【JavaScript】
https://qiita.com/kohki_takatama/items/4b99120da4f8e64cd791?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #JavaScript #Property #parameter #argument

Die #Gastronomie 🍽️ hat nach wie vor ein (Personal-) Problem, Gäste brauchen sie aber auch. Den meisten ist schon die Existenz von #OpenStreetMap nicht bekannt, geschweige denn die möglichen #Parameter https://wiki.openstreetmap.org/wiki/Tourism_tag_overview#Food_&_drinks

Insbesondere die steigende Anzahl apps 📱 https://wiki.openstreetmap.org/wiki/Comparison_of_Android_applications wäre ein Argument für zumindest den Eintrag der #Öffnungszeiten und des Links auf die #Webseite https://wiki.openstreetmap.org/wiki/Shop_suffix#Corresponding_prefixes

#Tourismus #BadenWürttemberg

Neue Studie grenzt Parameter zur #Expansion des Universums weiter ein

https://www.astropage.eu/2023/08/23/neue-studie-grenzt-parameter-zur-expansion-des-universums-weiter-ein/

#Standardkerze, #Entfernung, #Supernova, #Gammablitz, #Quasar, #Universum, #Parameter

Tails - Weak cryptographic parameters in LUKS1 https://tails.boum.org/security/argon2id/index.en.html #cryptsetup #parameter #dm-crypt #argon2id #weakness #crypto #pbkdf2 #tails #luks #kdf

Create Functions to Create SecureString SSM Parameter
~~
ACM.224 Creating a SecureString Parameter with the AWS CLI
~~
by Teri Radichel | May 27, 2023
#aws #securestring #parameter #cloud #data #security

https://medium.com/cloud-security/create-functions-to-create-securestring-ssm-parameter-802b418e4bfa

Create Functions to Create SecureString SSM Parameter
~~
ACM.224 Creating a SecureString Parameter with the AWS CLI
~~
by Teri Radichel | May 27, 2023
#aws #securestring #parameter #cloud #data #security

https://medium.com/cloud-security/create-functions-to-create-securestring-ssm-parameter-802b418e4bfa

@horenzienzwijgen @philipdutre ik volg de redenering, met 1 nuance:
SEO stemt inhoud en structuur af op een algoritme.
Prompt #Engineering gaat over aansturing van clusters van parameters.
Ik zou dit onderscheiden van Prompt #Crafting, wat gaat over de volledigheid en mate van detail die je invoert.
Crafting is een onderdeel van Engineering met een specifieker doel. #LLM #PARAMETER-TOYING

@horenzienzwijgen @philipdutre ik volg de redenering, met 1 nuance:
SEO stemt inhoud en structuur af op een algoritme.
Prompt #Engineering gaat over aansturing van clusters van parameters.
Ik zou dit onderscheiden van Prompt #Crafting, wat gaat over de volledigheid en mate van detail die je invoert.
Crafting is een onderdeel van Engineering met een specifieker doel. #LLM #PARAMETER-TOYING

Beifang:
Komplexität entsteht durch Erhöhung der Freiheitsgrade (#Parameter)

Die Reduktion auf genau 1 Freiheitsgrad (Zeit) erschafft einen lediglich komplizierten Ablauf.
Alle übrigen - ehemaligen Freiheitsgrade - wurden durch Festlegung (Entscheidungen) eliminiert.

Die so entstehenden Binärbäume können demnach nie einer tatsächlich existierenden Komplexität gerecht werden.

[Bash TIP of the day] : Bash supports kinda "named" parameters for function calls making your scripts much more readable.

Even the order of parameters does not matter anymore 🤩

#bash #tip #function #parameter

[Bash TIP of the day] : Bash supports kinda "named" parameters for function calls making your scripts much more readable.

Even the order of parameters does not matter anymore 🤩

#bash #tip #function #parameter

#Kurtosis was an important #parameter differentiating #size distributions, with #platykurtic distributions in #marls and #leptokurtic distributions in #limestones, suggesting that this parameter may reflect different degrees of #time #averaging. Most size #distributions were positively #skewed, but most strongly in marls. Complete #sampling led to #skewness values close to zero (#symmetrical distributions) and high kurtosis.

https://doi.org/10.2110/palo.2021.063

In case you missed our #kruize? 🛳️ All aboard! Watch the recording on YouTube 🎦https://op1.st/meetup-S02E03
#hpo #Hyper #Parameter #Optimization #docker #kubernetes #containers #openoperations #OperateFirst

#Verfassungsmäßigkeit von #Eingriff​en in #Grundrecht​e werden #geprüft entlang der #Frage nach #Verhältnismäßigkeit.

#Verhältnismäßig ist ein #Eingriff wenn er #entlang folgender #Parameter einer #Prüfung standhält:

#Zweck
#Geeignetheit
#Erforderlichkeit
#Angemessenheit = #Verhältnismäßigkeit im #eng​eren #Sinn.

https://dejure.org/gesetze/StPO/81a.html

#BVerfGE 16, 194 (200ff.); #Liquorentnahme; #Zulässigkeit #körperlich​er #Eingriff​e § 81a #StPO

👍 Neuer Service: Terminal-Tools

📌 Wer kennt das nicht, man möchte kurz mal was prüfen, aber dann fällt einem mal wieder nicht das passende Terminal-Tool bzw. Parameter ein.

#Terminal #Tools #Parameter #GNU #Linux #Befehle @kuketzblog

https://www.kuketz-blog.de/neuer-service-terminal-tools/