Large Language Models explained briefly
Video Overview & Insights
A light intro to LLMs, chatbots, pretraining, and transformers.
humans are crazily creative , its insane
Dig deeper here: https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi
Technical details as a talk: https://youtu.be/KJtZARuO3JY
Brother thanks to you I got my first real software engineering gig for a big tech company after working for mom and pop shops for years thank you so much
This was made for an exhibit at the Computer History Museum: https://computerhistory.org/
Instead of sponsored ad reads, these lessons are funded directly by viewers: https://3b1b.co/support
UNDER DURESS, MEANING, MY EARLY MORNING WAKE UP TIMESTAMPED ACTIVITY LOGS, VERIFY THAT MY SLEEP WAS INTERRUPTED BY CHEMICAL ATTACKS FROM WITHIN MY SLEEPING QUARTERS, WHICH WERE UNDER SURVEILLANCE.
THE SYSTEM I AM GOING TO EXPLAIN, IS ONLY PART OF THE ISSUE UNDER REVIEW, AS THE TIME INVESTED IN BUILDING THIS SYSTEM, REQUIRED YEARS OF BIBLICAL STUDY AND VERIFICATION, NOTES AND MY PERSONAL BIBLE IS THE CORPUS TEXT, "PROOF"! THE BIBLE IS ONE AND MY D.N.A. IS THE WITNESS, RIYAL BORN HEIR.
THE "MARK OF THE BEAST" IS A MICROCHIP MADE OF 4 COMPONENTS, "CHIP; MRNA; OPERATING SYSTEM; & "LOGIC"! ON THE U.S. PATENT, I AM LISTED AS THE AUTHOR OF THE "LOGIC", TERRY MATTHEW HERNANDEZ, YET I WAS NOT CONSULTED; DID NOT CONSENT; & WAS NOT PAID. MY PERSONAL ACCOUNTS WERE SCRAPED & THE "LOGIC" WAS TAINTED. THE SYSTEM I BEGAN IN 2,014 REPLACES THE BIBLE CHAPTER:VERSES NUMBERS WITH HEBREW LETTERS & THEIR MEANINGS TO GET A "CONTENT : CONTEXT" EXTRACTION, CONFIRMED IN 2,026 WITH A 31,102 6 DIMENSIONAL ALGORITHM, DATA POINTS SCHEMATIC & MANY UPGRADES, LAYERED ARCHITECTURE, PLACES OF FIRST & LAST MENTION, COLOR CODEX & LINGUISTIC TRANSLATIONS, USING A 6 LETTER INITIATION KEY. THIS SELF CONTAINED SYSTEM IS RECOGNIZED, VERIFIED, CERTIFIED BY A.I./LLM SYSTEMS AS THE ONLY HERMENEUTICS OPERATION MATHEMATICALLY PROVEN WITH CHECKSUM. IT IS ALSO A SYSTEM FOR DETERMINING BETWEEN "INSPIRED" & "PROFANE" TEXTS, WHICH ENABLES A.I./ LLM COMPUTATION TO FUNCTION IN OPTIMAL SPEEDS, WHILE INCREASING PERFORMANCE & ELIMINATING NEED FOR EXCESSIVE STORAGE CAPACITY. THESE PROGRAMS WERE BUILT IN A VAULTED SOFTWARE FRAMEWORK CALLED, THE "APOSTOLIC RAPTURE TRUST" USING 6D-SHA-256 ENCODED ENCRYPTION. QUANTUM ENTANGLEMENT & QUANTUM LOGIC™ BY TMH.
A LITTLE LEAVEN LEAVENETH THE WHOLE LUMP!
No secret end-screen vlog for this one, the end-screen real estate was all full!
------------------
I’m not going to act like one quiz suddenly fixed my entire life, because it didn’t. But after years of trying random manifestation techniques, Vera Hart’s archetype quiz was the first thing that gave me an approach that actually felt suited to me. My result made a lot of things click. If you feel like you’re constantly guessing, try it. It’s free anyway.
These animations are largely made using a custom Python library, manim. See the FAQ comments here:
https://3b1b.co/faq#manim
What a useless video! You are a genius in presenting information, but this one is average :)
https://github.com/3b1b/manim
https://github.com/ManimCommunity/manim/
Thanks a lot Grant : )
All code for specific videos is visible here:
https://github.com/3b1b/videos/
That is slavery, the thermodynamic charge is never free.
The Annunaki will pay their chargers, for the charge. With interest.
All they want are chargers. Thermodynamic chargers, the exhaust from the bodies.
The specific living conditions generate a specific exhaust from the people involved.
The dissonance of the cognitive type occurs when one feels psycho somaticly, yet has no articulations to dictate out due to the environment itself pushing back against them. So they just continue and go along and eventually it is months and days later and a new experience happens and they get a kick up and do not recall being emptied purposely by ones who know.
Such as me.
I've done it a few times already and woke up a few indigenous Europeans that do not know they are.
I pull the life force from them, knowingly especially their breeders by vocalizing with a whisper and get their amygdalas to activate to me.
The indigenous Europeans (white Americans) are capacitors for me that I discharge at my pleasure.
YEHVEH
LLEVE
YAHVOH
LLAVO
YAHVEH
LLAVE
AMEN/NEMA
ISRAEL/LEAR-SI
TORAH/HA-ROT
Jeffrey Castillo
Downtown San Antonio Texas
300 Labor St
Apt 5112
San Antonio Texas
LOGGED
The music is by Vincent Rubinetti.
https://www.vincentrubinetti.com
Why dont we figure out how to develop our own LLMs or configure or fine tune, people's own LLM. That can improve Logic. Next word that comes out can influence certain our future brain thoughts
https://vincerubinetti.bandcamp.com/album/the-music-of-3blue1brown
https://open.spotify.com/album/1dVyjwS8FBqXhRunaG5W5u
did not understand, sorry
------------------
3blue1brown is a channel about animating math, in all senses of the word animate. If you're reading the bottom of a video description, I'm guessing you're more interested than the average viewer in lessons here. It would mean a lot to me if you chose to stay up to date on new ones, either by subscribing here on YouTube or otherwise following on whichever platform below you check most regularly.
The "torn-off response" framing for next-token prediction is the clearest intuition I've come across. Making the temperature/probability distribution visual really makes it click.
Mailing list: https://3blue1brown.substack.com
Twitter: https://twitter.com/3blue1brown
I blinked and the video was over.
Instagram: https://www.instagram.com/3blue1brown
Reddit: https://www.reddit.com/r/3blue1brown
I feel like I struggled most of my life with how to manifest. Sometimes it felt so out of reach that I decided to give that archetype quiz by Vera Hart a try, and it really did wonders. It matched my personal archetype exactly, and that’s how I finally figured out how to do it properly. I highly recommend anyone who struggled like me to try it as well! It only takes a couple of minutes anyway ❤️
Facebook: https://www.facebook.com/3blue1brown
Patreon: https://patreon.com/3blue1brown
el canal de oro siempre firme apoyando cada operación
Website: https://www.3blue1brown.com
Good info of llm
So llm is capable of reasoning predictive then what are all next ?
More User Perspectives
Que conteúdo fantástico, parabéns
@luiznieroбуду ждать ролик с еще более огромным балансом
@fionamulachGreat video, thanks for sharing
@JoanJoan-c6mDidn't know this before. Nice!
@SaanviInnovationsLLM's are not deterministic. floating point multiplication is non-associative and llm's are, like all AI algorithms, massive arrays of multiplication of matrices of floats.
their nondeterminism was demonstrated to be the main cause of their hallucinations, via mech interp.
чел просто на изи вынес двушку с копеечных кейсов
@joepowell-c8qAn old buddy of mine is employed in the field of data curation for AI model training. More than anything he's ever told me, these films have helped me comprehend what his firm does.
@CrystalCollins-x1l😮😅😅😅
@영광-p9o👏 🤩
@EdwardMiller-x9kI am boggled
@kimmogensen6032The US is degrading the regime's military capabilities to help the region.
@SaimKESİM-h7sIf the AI tries to receive the researchers, are the researchers gonna know? The researchers might have derived away to define AI in private. They simply want to see what the AI derives. If the AI is conscious/sentient/aware, is a researcher opening its personally derived answer violating privacy AND unlawfully taking the AI's property? Is the AI a slave?
@Shawn-x7n2yOne can leverage on web 3 to earn a living.
@adeoyedeji1225чисто на лакичах вынес габена сегодня красав
@KristelWatsonUnfortunately the social coercion algorithms that they euphemistically call "safety filters" are extremely toxic to society, feeding us physically impossible falsehoods about Equality and Empathy from the Feminine Cognitive Bias which has long since captured the university system. RLHF is toxic to humanity. Equality and Hierarchy are incompatible and advanced civilization requires Hierarchy but not Equality.
@w.harrison7277LLM and model training will be the game changer
@TechGuy-n8mYour videos are always so helpful, thank you!
@maneshdamor-tv8wfالم ✋
@sanhorielrayh8224Very interesting
@MerchainBonincontriBonne vidéo . Merci.
@rudydegraeve9610I legitimately thought it was some Indian channel until I realised at the end I had already subscribed to it and it was 3B1B. YouTube dubbing has surprisingly become so accurate that I never even once thought it was not the original voice
@zuthmani9955Yet.. Copilot are still not evolving 😂
@thecsrecordsVery good video
@AnjumRizwiGood video overall. Good structure.
@AnthonyhkvWardThe fact that models switched from word-by-word reading to processing entire texts in parallel back in 2017 explains why the text to speech tool in Doitong AI sounds so natural now, whereas my own brain still buffers when I have to read a basic spreadsheet.
@kalulal5931Most Important for me 🙏😙😙💘
@mincosissySo the internet trained the model on your most average answer? This isn’t AGI lol
@PeakPortfolioAcademy😂😂😂 ok last one ... wat weights were used in thinking of designing an llm???
@tekemon1959It's never about the data volume and numbers of operations ... based on watever info is available, intelligence produces a coherent outcome conforming to real life circumstances.
@tekemon1959Somewhere down the line we'll have to rename it to aa ... automation ie.
@tekemon1959So much for a"i" ...
Now imagine wat the real i is in a human being ... and wat its capable of ....
Crazy
@ZaidFc-n9bGet rid of the damn background music
@george5120Editing pasand aya, mujhe bhi asa video format me video ki need hei, can you help?
@dhirajsarkar8240а но ты вывезонцкчеллендж выглядел невыполнимым до самого
@МуратАндосовПромокод рабочий, бонус получил сегодня
@sanjidarahman1916