Purvanshi Mehta (@purvanshi_mehta) 's Twitter Profile
Purvanshi Mehta

@purvanshi_mehta

Multimodal media transformations | co-founder @world_lica

ID: 963375832457719809

calendar_today13-02-2018 11:34:31

160 Tweet

428 Followers

1,1K Following

Purvanshi Mehta (@purvanshi_mehta) 's Twitter Profile Photo

One of the standout features of NotebookLM Podcasts that has gained significant popularity is its ability to turn any PDF or document into an audiobook. Interestingly, in February, we had a similar feature in Lica World , where users could generate engaging podcasts from documents

harsh (@harshsikka) 's Twitter Profile Photo

Excited to share our new paper "Benchmarking Vision, Language, & Action Models on Robotic Learning Tasks" We evaluate how well VLM & VLA models can control robots across 20 different real-world tasks. Experimental details, Models, Links, & Vision Below ā¬‡ļø This work was done

Excited to share our new paper "Benchmarking Vision, Language, & Action Models on Robotic Learning Tasks"

We evaluate how well VLM & VLA models can control robots across 20 different real-world tasks.

Experimental details, Models, Links, & Vision Below ā¬‡ļø

This work was done
Lica World (@world_lica) 's Twitter Profile Photo

šŸ‘‹ Meet Lica. Chat to craft product demos, design reviews, strategy presentations, and more. From tweaking styles to refining scripts and voiceovers, she makes video creation super easy. Skip the long meetings and endless emails—send a video instead!

AI User Group (@aiusergroup) 's Twitter Profile Photo

Check out this short by Purvanshi Mehta on "Turning a Husky’s Life Into a Blockbuster with AI! šŸ¶šŸŽ„". Watch the full replay here: youtu.be/UhcOajHrTYc #AI #AIForDesigners Lica World

Ryan Marten (@ryanmart3n) 's Twitter Profile Photo

Announcing the Open Thoughts project. We are building the best reasoning datasets out in the open. Building off our work with Stratos, today we are releasing OpenThoughts-114k and OpenThinker-7B.

Announcing the Open Thoughts project. We are building the best reasoning datasets out in the open.

Building off our work with Stratos, today we are releasing OpenThoughts-114k and OpenThinker-7B.
Purvanshi Mehta (@purvanshi_mehta) 's Twitter Profile Photo

Anthropic’s ASL-3 disclosure is a step in the right direction, but it also makes one thing clear: We’re relying on voluntary guardrails for systems that won’t stay voluntary for long. ASL-3 refers to ā€œAI Safety Level 3ā€ — a classification Anthropic uses for models with dangerous

Dylan Itzikowitz (@ditzikow) 's Twitter Profile Photo

Interviewing fellow šŸ‡ØšŸ‡¦'s is my favourite pastime. David Baszucki came by South Park Commons to share his story building Roblox into a 100m daily active user platform! Check out the full conversation in the links below!

Purvanshi Mehta (@purvanshi_mehta) 's Twitter Profile Photo

Hiring engineers today looks different. It’s not just about writing clean code — it’s about speed, systems thinking, and knowing how to work with tools like Cursor and Copilot. In an interview today, I realized how much my expectations have shifted. And I’m hearing the same

Hiring engineers today looks different.

It’s not just about writing clean code — it’s about speed, systems thinking, and knowing how to work with tools like Cursor and Copilot.

In an interview today, I realized how much my expectations have shifted. And I’m hearing the same
Purvanshi Mehta (@purvanshi_mehta) 's Twitter Profile Photo

One of the technical challenges while building foundational models for graphic design is SVG generation. SVGs aren’t just text — they’re structured graphics with layers, transformations, clip paths, gradients, and interdependent properties. Training autoregressive models to

One of the technical challenges while building foundational models for graphic design is SVG generation.

SVGs aren’t just text — they’re structured graphics with layers, transformations, clip paths, gradients, and interdependent properties. Training autoregressive models to