Parisa Rashidi
@parisarashidi.bsky.social
1.9K followers 2.2K following 110 posts
Scientist, Medical AI Researcher, Humanist Founding Co-director of @UF_IC3 Center and director of @UFiHealLab #AI4Health Opinions my own. She/her.
Posts Media Videos Starter Packs
Reposted by Parisa Rashidi
parisarashidi.bsky.social
Isn't it wonderful that AI is auto-generating my commit messages? Took away all the guilt of not including enough details!
Reposted by Parisa Rashidi
carldoersch.bsky.social
We're very excited to introduce TAPNext: a model that sets a new state-of-art for Tracking Any Point in videos, by formulating the task as Next Token Prediction. For more, see: tap-next.github.io
Reposted by Parisa Rashidi
stephanjanssen.be
Google just open sourced a Python Agent Development Kit @ github.com/google/adk-p...
Native Streaming and #MCP support out of the box 👀
github.com/google/adk-s...
Reposted by Parisa Rashidi
erictopol.bsky.social
2 important papers @nature.com today indicate the potential for A.I. to improve diagnostic accuracy and overall clinical performance for patient management and interactions
nature.com/articles/s41...
Reposted by Parisa Rashidi
Reposted by Parisa Rashidi
asheeshksi.bsky.social
These cuts would be catastrophic for the humanities and for communities around the country - you can contact your reps to oppose these cuts here: p2a.co/DdtlGIT We should be increasing resources for humanities scholarship and programs, not cutting them! 🗃️
parisarashidi.bsky.social
It is hard enough to recruit one qualified researcher. Even if the next administration tries to reinstate all these offices, it would be a tall order to find and recruit all the qualified researchers again, they already have moved on.
Reposted by Parisa Rashidi
deniswirtz.bsky.social
This is catastrophic for pancreatic cancer patients
Reposted by Parisa Rashidi
jeffdean.bsky.social
🥁Introducing Gemini 2.5, our most intelligent model with impressive capabilities in advanced reasoning and coding.

Now integrating thinking capabilities, 2.5 Pro Experimental is our most performant Gemini model yet. It’s #1 on the LM Arena leaderboard. 🥇
Reposted by Parisa Rashidi
montfort.me
Please pay attention to what is going on with the deportation mess. Individual rights are being trampled so ICE can make quota.
joycewhitevance.bsky.social
1/ For many Americans, deporting violent gang members will sound like a good idea, constitutional niceties be damned. They are the perfect group of people for a government test case. joycevance.substack.com/p/deportatio...
Deportations: It's not where it starts, it's where it ends.
Just ahead of the hearing scheduled for 5 p.m.
joycevance.substack.com
parisarashidi.bsky.social
Our goal with NeuRL is to make AI education accessible to everyone. We’re expanding it into a comprehensive AI education platform and exploring integrations with other institutions. Get in touch!
parisarashidi.bsky.social
It is a custom neural network and RL library in OpenGL Shading Language (GLSL) to enable GPU-accelerated training—making real-time agent updates possible without deep learning libraries.
parisarashidi.bsky.social
NeuRL solves this by:
✅ Running entirely in the browser, with no installations required
✅ Providing real-time visualization of neural networks and agent behaviors
✅ Supporting multiple RL algorithms and interactive tuning
parisarashidi.bsky.social
Why did we create NeuRL?
Traditional AI education often relies on code-heavy tools and high-performance computing resources. This excludes many students from experimenting with RL concepts interactively.
parisarashidi.bsky.social
NeuRL is a free and accessible tool that enables students to interactively explore RL and neural networks using 3D WebGL environments—no coding required.
parisarashidi.bsky.social
🚀 Our latest paper presented at #SIGCSE2025. @snsie.bsky.social, my postdoc, has developed NeuRL, a no-code, web-based platform for real-time exploring of neural networks and reinforcement learning (RL).

🔗 neu-rl.netlify.app
📜 paper: dl.acm.org/doi/abs/10.1...

Here's a brief overview 🧵👇
Reposted by Parisa Rashidi
emollick.bsky.social
I regret to announce that the meme Turing Test has been passed

LLMs produce funnier memes than the average human, as judged by humans. Humans working with AI get no boost (a finding that is coming up often in AI-creativity work) The best human memers still beat AI, however. arxiv.org/abs/2501.11433
Reposted by Parisa Rashidi
docsirianni.bsky.social
This is horrible to post, but I may as well post it. We are essentially shutting down research operations in my group, which is focused on treatments for pediatric brain cancer. I’m a well funded investigator, and there’s no choice. Science can’t function without the stability of NIH
Reposted by Parisa Rashidi
erictopol.bsky.social
A relentless attack on universities proceeds