Loading video...

Video Failed to Load

Go Home

In our blog post, you can interactively explore the reasoning effect of token cues and RL on base models! 📎 paper: ✍ blog: 💻 code:

SophieLWang's profile picture

Sophie Wang

2,167 subscribers

20,345 views • 3 days ago •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Can “chicken” make a base model reason? 🐔 Yes! Why? With the right first tokens, base models can match RL in reasoning. We trace this to learned training data associations and use a simple data edit to make “chicken” elicit reasoning too. blog:
0:10

Can “chicken” make a base model reason? 🐔 Yes! Why? With the right first tokens, base models can match RL in reasoning. We trace this to learned training data associations and use a simple data edit to make “chicken” elicit reasoning too. blog:

Sophie Wang

131,416 views • 4 days ago

I've updated my blog post to walk through the remaining technical details of our Surface Winding Numbers algorithm: now the calculus of the algorithm is explained a bit more in detail. The post, paper, code, etc. is all here:
0:10

I've updated my blog post to walk through the remaining technical details of our Surface Winding Numbers algorithm: now the calculus of the algorithm is explained a bit more in detail. The post, paper, code, etc. is all here:

Nicole Feng

37,032 views • 1 year ago

In his first blog post on Jeff Smith shares firsthand knowledge and tips on maintaining a healthy relationship with the digital world. To read the full blog:
0:19

In his first blog post on Jeff Smith shares firsthand knowledge and tips on maintaining a healthy relationship with the digital world. To read the full blog:

Therapy.ai

39,286 views • 2 years ago

It's possible to run custom code on the Nintendo Alarmo via USB - without opening it up! More details in the blog post here:
0:17

It's possible to run custom code on the Nintendo Alarmo via USB - without opening it up! More details in the blog post here:

Gary

582,152 views • 1 year ago

🚀Product Announcement: Built on the Infrastructure layer DePIN of Shieldeum, is designed to serve as both an AI Token Index and AI Agent Creator. Learn more in our detailed blog:
0:11

🚀Product Announcement: Built on the Infrastructure layer DePIN of Shieldeum, is designed to serve as both an AI Token Index and AI Agent Creator. Learn more in our detailed blog:

Shieldeum VPN

157,352 views • 1 year ago

Paper blog is here! We explore retro print aesthetics with our new HalftoneCMYK shader. I break down the tech tricks behind it, and Agu Seguí shows how to use it Plenty of interactive illustrations (and plenty of banana pictures). Enjoy! read: paper app: shader docs:
0:10

Paper blog is here! We explore retro print aesthetics with our new HalftoneCMYK shader. I break down the tech tricks behind it, and Agu Seguí shows how to use it Plenty of interactive illustrations (and plenty of banana pictures). Enjoy! read: paper app: shader docs:

Ksenia Kondrashova

39,345 views • 8 months ago

If you want to see a technical breakdown of the 35mm website, I wrote a blog post on Codrops where I detail the Three.js/shaders code ! It's live on Codrops🔗⬇️
0:19

If you want to see a technical breakdown of the 35mm website, I wrote a blog post on Codrops where I detail the Three.js/shaders code ! It's live on Codrops🔗⬇️

chakib

15,126 views • 1 year ago

We’re continuing to modernize our IDEs and have refreshed something you interact with constantly — the editor caret and selection! Learn more in this blog post 👇
0:12

We’re continuing to modernize our IDEs and have refreshed something you interact with constantly — the editor caret and selection! Learn more in this blog post 👇

IntelliJ IDEA, a JetBrains IDE

27,865 views • 7 months ago

Qubic x OKx Campaign. Happening right now and offering some good rewards! See the details in our blog post⏬
0:11

Qubic x OKx Campaign. Happening right now and offering some good rewards! See the details in our blog post⏬

Qubic

31,205 views • 1 year ago

We are excited to launch our two models Pharia-1-LLM-7B-control and Pharia-1-LLM-7B-control-aligned. Both models and the code used to train them are now publicly available and open-sourced for non-commercial research and educational use. Read our model blog post here: Learn more about our open-source codebase Scaling: #writtenbyalephalpha
0:13

We are excited to launch our two models Pharia-1-LLM-7B-control and Pharia-1-LLM-7B-control-aligned. Both models and the code used to train them are now publicly available and open-sourced for non-commercial research and educational use. Read our model blog post here: Learn more about our open-source codebase Scaling: #writtenbyalephalpha

Aleph Alpha

44,380 views • 2 years ago

35.9 tok/sec on Windows 98 🤯 This is a 260K LLM with Llama-architecture. We also tried out larger models. Results in the blog post.
0:17

35.9 tok/sec on Windows 98 🤯 This is a 260K LLM with Llama-architecture. We also tried out larger models. Results in the blog post.

Alex Cheema

569,400 views • 1 year ago

Here's `notes-to-blog`, an insanely powerful HyperWrite tool that turns any set of notes into a full blog post. And it doesn't sound like a robot wrote it. You can paste in: - interview transcripts - podcast notes - your own research The final output is fantastic. Try it:
0:16

Here's `notes-to-blog`, an insanely powerful HyperWrite tool that turns any set of notes into a full blog post. And it doesn't sound like a robot wrote it. You can paste in: - interview transcripts - podcast notes - your own research The final output is fantastic. Try it:

Matt Shumer

41,739 views • 2 years ago

We believe in rewarding our early believers and supporters. Early contributors will be rewarded for helping us build our community. Don’t miss out! 🪂👀 Your first steps: 1. Join our Discord 2. Follow us on X Read all about are points program in our blog post 👇🏻 Read more on our blog: Discord: XSY DApp:
0:19

We believe in rewarding our early believers and supporters. Early contributors will be rewarded for helping us build our community. Don’t miss out! 🪂👀 Your first steps: 1. Join our Discord 2. Follow us on X Read all about are points program in our blog post 👇🏻 Read more on our blog: Discord: XSY DApp:

YieldPoint

71,135 views • 1 year ago

We know the threats of viruses and bacteria, but what about fungi? 🍄🔍 Fungi can be part of our normal flora, but one stands out for its resilience. Our blog post explains why we are keeping a close watch on Candidozyma auris in particular. 🔗
0:10

We know the threats of viruses and bacteria, but what about fungi? 🍄🔍 Fungi can be part of our normal flora, but one stands out for its resilience. Our blog post explains why we are keeping a close watch on Candidozyma auris in particular. 🔗

UK Health Security Agency

15,052 views • 9 months ago

Blog filtering in Framer ✨ ✅ No override code ✅Uses component logic Comment "Framer" and I'll send you the remix link (must be following so I can DM you) ❤️
0:15

Blog filtering in Framer ✨ ✅ No override code ✅Uses component logic Comment "Framer" and I'll send you the remix link (must be following so I can DM you) ❤️

Ryan Hayward

17,777 views • 3 years ago

TWO IN TWO. Starc gets Bavuma and the Australian's are on a team hat-trick! 📺 Watch #AUSvSA on Ch. 501 or stream via Kayo Sports ✍ BLOG 📲 MATCH CENTRE
0:13

TWO IN TWO. Starc gets Bavuma and the Australian's are on a team hat-trick! 📺 Watch #AUSvSA on Ch. 501 or stream via Kayo Sports ✍ BLOG 📲 MATCH CENTRE

Fox Cricket

17,768 views • 3 years ago

AI video is here and it's powerful. But it can be tricky to get the exact video you want. We've made it easier by adding video models to the Replicate playground. You can prompt for videos directly, or generate images and then animate them. Read more on on our blog. Link below
0:12

AI video is here and it's powerful. But it can be tricky to get the exact video you want. We've made it easier by adding video models to the Replicate playground. You can prompt for videos directly, or generate images and then animate them. Read more on on our blog. Link below

Replicate

29,168 views • 1 year ago

✨🧠 Tribe v2, our latest model of human brain responses to sound, sight and language can now be (partly) explored on your phone📱: ▶️demo: 📄paper: 💻code:
0:11

✨🧠 Tribe v2, our latest model of human brain responses to sound, sight and language can now be (partly) explored on your phone📱: ▶️demo: 📄paper: 💻code:

Jean-Rémi King

38,521 views • 5 months ago