New Stability AI blog post on our MindEye2 work showing state-of-the-art reconstructions of perception from fMRI brain activity.

Our approach shows how high-quality reconstructions can be obtained w/ just a fraction of training data for a new subject!

stability.ai/news/mindeye2-…

Today would have been my grandma's 91st birthday.

Unfortunately she passed away a few weeks ago. She's now celebrating her first heavenly birthday with my grandfather.

My grandma, Dr. Thankam Mathew, always had a strong presence in our lives from the very beginning. She was…

Today would have been my grandma's 91st birthday.

Unfortunately she passed away a few weeks ago. She's now celebrating her first heavenly birthday with my grandfather.

My grandma, Dr. Thankam Mathew, always had a strong presence in our lives from the very beginning. She was…

This week so far:
• 'Majorly improved' GPT-4 roll out
• Udio samples dropping on the timeline
• Gemini 1.5 Pro wider free access w/ audio
understanding, unlimited files, etc.
• Mistral 8x22B weights release
• sooo much more

yeah... it's time to pull out this meme again

This week so far:
• 'Majorly improved' GPT-4 roll out
• Udio samples dropping on the timeline
• Gemini 1.5 Pro wider free access w/ audio
understanding, unlimited files, etc.
• Mistral 8x22B weights release
• sooo much more

yeah... it's time to pull out this meme again

Eagle and Finch: RWKV with Matrix-Valued States and Dynamic Recurrence

arxiv.org/abs/2404.05892
code: github.com/RWKV/RWKV-LM

Describes RWKV-5 and RWKV-6 which significantly improve over RWKV-4, scaled up to 7.5b and 3.1b billion multilingual models respectively. Completely…

Total solar eclipse at IU Bloomington #Eclipse2024 taken with my smartphone 😅

It's getting cooler! Almost total eclipse! #Eclipse2024

At IU Bloomington watching the total solar eclipse!

lol

One of our co-founders recently met with Dr. Shane Oberloier to talk about our Portality platform and he made an awesome video about it!

youtube.com/watch?v=z5mqhh…

Thanks for taking the time to chat with us and try out our platform, Shane!

OpenAI publicly announces their Voice Engine, which allows voice cloning from 15 seconds of audio.

openai.com/blog/navigatin…

Originally developed in late 2022, they have tested it with a variety of trusted partners. Some demo samples are shared in the blog post. They have no…

X.AI announces Grok-1.5

x.ai/blog/grok-1.5

'Grok-1.5 achieved a 50.6% score on the MATH benchmark and a 90% score on the GSM8K benchmark, two math benchmarks covering a wide range of grade school to high school competition problems.'

Better than Mistral…

.Paul Scotti and I will talk about MindEye2 tomorrow at 📅 ThursdAI, come join us!

Are you ever fed up with the lack of sufficiently fine-grained control in T2I diffusion models?
It turns out diffusion models already knew how to do it, we just have to learn how to leverage this capability.
Project page: compvis.github.io/attribute-cont… (1/n)