Your curated collection of saved posts and media

Showing 32 posts Β· last 7 days Β· newest first
_
_akhaliq
@_akhaliq
πŸ“…
Dec 11, 2025
113d ago
πŸ†”69255899

discuss: https://t.co/uJHaaQkg5G

Media 1
πŸ–ΌοΈ Media
S
SergioPaniego
@SergioPaniego
πŸ“…
Dec 10, 2025
114d ago
πŸ†”85947968

We just released TRL v0.26.0! It comes packed with updates: > Agent training with tools in GRPO > New CISPO & SAPO losses > Reasoning rewards > vLLM quantization in colocate mode > Dataset shuffling in SFT > Lots of NEW examples > Tons of fixes and documentation improvements https://t.co/Vt3dmI1sLU

Media 1
πŸ–ΌοΈ Media
L
LoubnaBenAllal1
@LoubnaBenAllal1
πŸ“…
Dec 11, 2025
113d ago
πŸ†”74888249

Sharing the slides from a talk I gave this week on bridging the gap between research experiments and building production-ready models, based on our recent Smol Training Playbook. https://t.co/RmG53PytMv

Media 1
πŸ–ΌοΈ Media
E
essential_ai
@essential_ai
πŸ“…
Dec 10, 2025
114d ago
πŸ†”35110383

We are now the #1 trending text-gen <256B size model on HuggingFace!! https://t.co/SyoOHjWfvH

Media 1
πŸ–ΌοΈ Media
πŸ”huggingface retweeted
E
Essential AI
@essential_ai
πŸ“…
Dec 10, 2025
114d ago
πŸ†”35110383

We are now the #1 trending text-gen <256B size model on HuggingFace!! https://t.co/SyoOHjWfvH

Media 1Media 2
❀️69
likes
πŸ”9
retweets
πŸ–ΌοΈ Media
O
OpenBMB
@OpenBMB
πŸ“…
Dec 10, 2025
114d ago
πŸ†”77921144

πŸ”₯ Ultra-FineWeb-en-v1.4 is coming! 2.2T tokens fully open-sourced! The core training fuel for MiniCPM4 / 4.1, fully updated based on FineWeb v1.4.0: πŸ†• What's New 1️⃣ Fresher Data: Added CommonCrawl snapshots from Apr 2024 - Jun 2025 to capture the latest world knowledge. 2️⃣ Easier Access: CC Dump Slices are here! No need to download the entire massive dataset anymore, fetch exactly what you need seamlessly. ⚑ Highlights & Performance - Efficient Verification: Efficient Verification Strategy: Reduces data verification cost by 90% - High-Efficiency Filtering Pipeline: Optimizes selection of both positive and negative samples - Performance Gains: +3.613/+1.331 (Eng) & +1.98/+0.61 (Chn) vs. FineWeb/FineWeb-edu & Chinese FineWeb-edu-v2. Still high-quality cleaning. Still true to the open-source spirit. Welcome to download and test! πŸš€ πŸ”— Resources πŸ€— Dataset: https://t.co/KluL5t2kUn πŸ“„ Paper: https://t.co/Kg9LLUqZgB 🧩 Classifier:https://t.co/oUfxrN6AmP πŸ€– MiniCPM4:https://t.co/IQ82jD1PTi #UltraFineWeb #MiniCPM4 #AI #LLM #OpenBMB #UltraData

Media 1Media 2
+1 more
πŸ–ΌοΈ Media
G
ggerganov
@ggerganov
πŸ“…
Dec 10, 2025
114d ago
πŸ†”07435615

> llama-cli -hf org/model

Media 1
πŸ–ΌοΈ Media
πŸ”huggingface retweeted
G
Georgi Gerganov
@ggerganov
πŸ“…
Dec 10, 2025
114d ago
πŸ†”07435615

> llama-cli -hf org/model

Media 1Media 2
❀️470
likes
πŸ”56
retweets
πŸ–ΌοΈ Media
V
victormustar
@victormustar
πŸ“…
Dec 10, 2025
114d ago
πŸ†”66967769

llama.cpp gets a new CLI (tested it and it's πŸ”₯) https://t.co/XKVicocKGC

Media 1
πŸ–ΌοΈ Media
πŸ”huggingface retweeted
V
Victor M
@victormustar
πŸ“…
Dec 10, 2025
114d ago
πŸ†”66967769

llama.cpp gets a new CLI (tested it and it's πŸ”₯) https://t.co/XKVicocKGC

Media 1Media 2
❀️170
likes
πŸ”16
retweets
πŸ–ΌοΈ Media
C
ClementDelangue
@ClementDelangue
πŸ“…
Dec 11, 2025
113d ago
πŸ†”16139534

Even if you don't have a reachy mini (yet!), you can now creates apps thanks to our SDK, API and simulation and share them with the community. If you create simple apps in the coming days, I'll try them on my mini and share a video of them (+ you'll probably get good visibility as we're shipping a large number soon). Some ideas I had: - "What is love" Reachy mini plays "what is love?" and do the classic Jim Carrey head move - "Metronome app": Reachy's antennas turn into a metronome - "Relax app": Reachy plays relaxing music and does some calm zen moves - "Magic 8-Ball" Reachy answers a simple yes/no question by nodding or shaking its head based on a random outcome. - "Peek-a-Boo": Reachy stays hidden until an object (like a hand) gets close, then quickly pops its head up. - "Bless you": Reachy mini says "bless you" when you cough - "Take a picture": Reachy mini takes a picture of you - "Read": Reachy mini reads the paper you show him (using OCR) - "Face-tracking app" simple face-tracking app - Reachy follows your face as you move around - "Who's speaking": Reachy tries to detect who's speaking and turn/concentrate on them - "Describe the room": Reachy mini scans the room and describes it - "Describe what's on my screen": Reachy detects your laptop screen and describes what's on it - "Describe the object" you show an object, Reachy tells you what it is - "Where are?" Where are my keys, Reachy scans the room and point to where they are - "Dance app": Reachy mini does simple dances - "Translation app": you say something in English, Reachy repeats it in French So many ideas haha

Media 1
πŸ–ΌοΈ Media
A
ashVaswani
@ashVaswani
πŸ“…
Dec 11, 2025
113d ago
πŸ†”36509796

Rnj-1-Instruct is now the #1 trending text generation model on HF! https://t.co/Gt5WGq9vLp

Media 1
πŸ–ΌοΈ Media
πŸ”huggingface retweeted
A
Ashish Vaswani
@ashVaswani
πŸ“…
Dec 11, 2025
113d ago
πŸ†”36509796

Rnj-1-Instruct is now the #1 trending text generation model on HF! https://t.co/Gt5WGq9vLp

Media 1Media 2
❀️98
likes
πŸ”11
retweets
πŸ–ΌοΈ Media
A
AndrewYNg
@AndrewYNg
πŸ“…
Dec 11, 2025
113d ago
πŸ†”59770795

Sharing a fun recipe for building a highly autonomous, moderately capable, and very UNreliable agent using the open source aisuite package that Rohit Prasad and I have been working on. With a few lines of code, you can give a frontier LLM a tool (like disk access or web search), prompt it with a high-level task (such as creating a snake game and saving as an HTML file, or carrying out deep research), and let the LLM loose and see what it does. Example in image. Caveat: This is not how practical agents are built today, since most need much more scaffolding (see my Agentic AI course to learn more), but is still interesting to experiment with. Longer write-up here: https://t.co/BdS8tGhnIy

Media 1Media 2
πŸ–ΌοΈ Media
A
arcprize
@arcprize
πŸ“…
Dec 11, 2025
113d ago
πŸ†”45547795

A year ago, we verified a preview of an unreleased version of @OpenAI o3 (High) that scored 88% on ARC-AGI-1 at est. $4.5k/task Today, we’ve verified a new GPT-5.2 Pro (X-High) SOTA score of 90.5% at $11.64/task This represents a ~390X efficiency improvement in one year https://t.co/9T47FdZ5Ry

Media 1
πŸ–ΌοΈ Media
M
ManusAI
@ManusAI
πŸ“…
Dec 11, 2025
113d ago
πŸ†”16573331

A website that doesn't rank is a missed opportunity. We’ve added a complete SEO engine to Manus Web App so you can stop worrying about config and start getting traffic. Turn your site into a growth engine today. https://t.co/c8wZzgURBM https://t.co/EP7ITH4YV0

Media 1Media 2
πŸ–ΌοΈ Media
M
MattZeitlin
@MattZeitlin
πŸ“…
Dec 11, 2025
113d ago
πŸ†”90566503

the Kalshi bets for Time Person of the Year being "AI" resolved to No because the actual winner was a bunch of AI executives they called "Architects of AI" https://t.co/TwOipqmjt1

Media 1
πŸ–ΌοΈ Media
T
TheWapplehouse
@TheWapplehouse
πŸ“…
Dec 11, 2025
113d ago
πŸ†”68087098

I’ve watched this 200 times https://t.co/JhwGXUMsI6

πŸ–ΌοΈ Media
πŸ”youwouldntpost retweeted
T
Kristi Yamaguccimane
@TheWapplehouse
πŸ“…
Dec 11, 2025
113d ago
πŸ†”68087098

I’ve watched this 200 times https://t.co/JhwGXUMsI6

❀️5,924
likes
πŸ”535
retweets
πŸ–ΌοΈ Media
J
jonathanmb32
@jonathanmb32
πŸ“…
Dec 11, 2025
113d ago
πŸ†”96490104

https://t.co/ydnVN9sHcw

@maxwelltani β€’ Thu Dec 11 15:39

American Canto sold 1165 physical copies in its first week, per bookscan.

Media 1
πŸ–ΌοΈ Media
πŸ”youwouldntpost retweeted
J
Jonathan
@jonathanmb32
πŸ“…
Dec 11, 2025
113d ago
πŸ†”96490104

https://t.co/ydnVN9sHcw

Media 1Media 2
❀️14
likes
πŸ”1
retweets
πŸ–ΌοΈ Media
W
willsommer
@willsommer
πŸ“…
Dec 11, 2025
113d ago
πŸ†”37905176

Insane story about @JackPosobiec threatening a reporter over a Pentagon official's apparent cuckolding Goodreads account. https://t.co/FeQLirpVhV

Media 1
πŸ–ΌοΈ Media
πŸ”youwouldntpost retweeted
Y
youwouldntpost
@youwouldntpost
πŸ“…
Dec 11, 2025
113d ago
πŸ†”95932320

Media 1Media 2
πŸ–ΌοΈ Media
G
github
@github
πŸ“…
Dec 11, 2025
113d ago
πŸ†”45118342

.@OpenAI’s GPT-5.2 is now rolling out in public preview in GitHub Copilot. This model is focused on long context and front-end UI generation. Try it out in @code ⬇️ https://t.co/0QSRnzTq5z

Media 1
πŸ–ΌοΈ Media
J
JohnLeFevre
@JohnLeFevre
πŸ“…
Dec 11, 2025
113d ago
πŸ†”28615319

chilling "kill all white people." https://t.co/499gfozgJq

πŸ–ΌοΈ Media
M
MarioNawfal
@MarioNawfal
πŸ“…
Dec 11, 2025
113d ago
πŸ†”47142219

πŸš¨πŸ‡ΈπŸ‡» BREAKING: ELON AND EL SALVADOR LAUNCH WORLD’S 1ST NATIONWIDE AI TUTORING PROGRAM! Elon and El Salvador’s President Bukele are shaking up the world with a game-changing move. Elon tweeted, β€œGrok will be used nationwide by El Salvador for personalized education!” While xAI confirmed, β€œGrok for Education: xAI is thrilled to announce a partnership with El Salvador and Nayib Bukele to bring personalized Grok tutoring to every public-school student in the country, over 1 million children. The world’s first nationwide AI tutor program.” While most of Latin America clings to outdated classrooms, chalkboards, overcrowded desks, and teachers stretched thin, El Salvador and Elon are diving into the future with AI. This isn’t just tech flexing; it’s personal. Imagine a kid in San Salvador getting a custom tutor via Grok, learning at their own pace, while neighbors in Guatemala or Honduras still memorize from dusty textbooks. Bukele’s betting big on this, syncing with Elon’s vision to leapfrog education, potentially lifting 1M kids out of poverty with skills for tomorrow’s jobs. Latin peers like Mexico (stuck in teacher strikes) or Brazil (underfunded schools) look stuck in the past, proving El Salvador’s bold step into a bright future, while others wallow in gray stagnation. This could spark a regional race, or leave laggards in the dust! Source: @elonmusk, @xai, @nayibbukele

@elonmusk β€’ Thu Dec 11 15:01

Grok will be used nationwide by El Salvador for personalized education!

Media 1Media 2
+1 more
πŸ–ΌοΈ Media
O
OpenAI
@OpenAI
πŸ“…
Dec 11, 2025
113d ago
πŸ†”62668275

GPT-5.2 Thinking evals https://t.co/Kcnz3ZIwye

Media 1Media 2
πŸ–ΌοΈ Media
O
OpenAI
@OpenAI
πŸ“…
Dec 11, 2025
113d ago
πŸ†”51212167

On GDPval, an eval measuring well-specified knowledge work tasks across 44 occupations, GPT-5.2 Thinking is our first model that performs at a human expert level. These tasks include making presentations, spreadsheets, and other artifacts. https://t.co/vyKSJrYLgG

Media 1Media 2
πŸ–ΌοΈ Media
O
OpenAI
@OpenAI
πŸ“…
Dec 11, 2025
113d ago
πŸ†”53694367

GPT-5.2 Thinking Raises the bar for professional work: - State-of-the-art long-context reasoning - Major improvements in spreadsheet creation, analysis, and formatting - Early gains in slideshow creation

Media 1Media 2
πŸ–ΌοΈ Media
J
JasonMiller
@JasonMiller
πŸ“…
Dec 10, 2025
114d ago
πŸ†”67644557

INSANE!!! A real-life β€œDavid vs. Goliath.” Do we really want ONE company having this much power to dictate what types of shows we watch??? β€œNetflix's equity value is nearly $40 billion greater than that of all other major media companies and theatrical exhibitors combined.” https://t.co/vbBtYIju1R

Media 1
πŸ–ΌοΈ Media
πŸ”johnrobinsn retweeted
_
AK
@_akhaliq
πŸ“…
Dec 11, 2025
113d ago
πŸ†”95901719

OmniPSD Layered PSD Generation with Diffusion Transformer https://t.co/IQnUyjCsZb

❀️57
likes
πŸ”13
retweets
πŸ–ΌοΈ Media
N
native_info
@native_info
πŸ“…
Nov 29, 2016
3412d ago
πŸ†”22682368

∴ γƒγ‚€γƒ†γ‚£γƒ–γƒ­γ‚°ζ›΄ζ–°γ—γΎγ—γŸγƒΌγ€‚ 硢賛受注中!γͺγ‚“γ¨γ€ŒδΊœηΎŽγ€γ‚’ζγ„γ¦γγ γ•γ£γŸHisasiε…ˆη”Ÿγ‹γ‚‰γ‚³γƒ‘γƒ³γƒˆγ‚’ι ‚γγΎγ—γŸγγγγγοΌοΌγ‚γ‚γ‚γ‚γ‚γ‚ŠγŒγ¨γ†γ”γ–γ„γΎγ™β€¦οΌοΌοΌ(Β΄οΌ›Ο‰οΌ›`) https://t.co/KBfiqvxnxt https://t.co/iWr6eYtIMO

Media 1Media 2
+3 more
πŸ–ΌοΈ Media