Increase access & opportunity for every human. Prev: Llama @ Meta, General Catalyst, Facebook AI, Mozilla.
Home
Joined April 2008
- Tweets4.3K
- Following724
- Followers910
- Likes1.9K
Awesome to see BioStack come out of stealth. High quality, longitudinal patient trajectories across diverse demographics are foundational to train agents that improve health outcomes. Reach out to @sanatmishra7 & team if you're building in this space!
Today we're sharing a major milestone for BioStack:
We’ve raised $9.5 million from @peakxvpartners, @EmCollective, @DCVC, @AforeVC, @VerdictCap and @ycombinator among several others to build the data foundation for the next generation of healthcare AI.
🧵
Incredibly proud of @manohar_paluri and team for taking on this challenge. I've worked with him for more than a decade and can't think of a better person to go after this mission. Rooting for team warp!
So excited to be part of Warp founding team and partnering with Kevin, Sol & DJ on building out Warp to accelerate scientific discovery starting with helping drug discovery! Checkout warp.xyz
Hey @cathaypacific - I have been trying to get a hold of your customer support across all channels for the past several hours to no avail. Can someone reply to my WhatsApp or X messages?
Ragavan retweeted
Meta COOKED! Llama 4 is out! Llama 4 Maverick (402B) and Scout (109B) - natively multimodal, multilingual and scaled to 10 MILLION context! BEATS DeepSeek v3🔥
Llama 4 Maverick:
> 17B active parameters, 128 experts, 400B total parameters > Beats GPT-4o & Gemini 2.0 Flash, competitive with DeepSeek v3 at half the active parameters > 1417 ELO on LMArena (chat performance). > Optimized for image understanding, reasoning, and multilingual tasks
Llama 4 Scout:
> 17B active parameters, 16 experts, 109B total parameters
> Best-in-class multimodal model for its size, fits on a single H100 GPU (with Int4 quantization)
> 10M token context window
> Outperforms Gemma 3, Gemini 2.0 Flash-Lite, Mistral 3.1 on benchmarks
Architecture & Innovations
> Mixture-of-Experts (MoE):
First natively multimodal Llama models with MoE
> Llama 4 Maverick: 128 experts, shared expert + routed experts for better efficiency.
Native Multimodality & Early Fusion:
> Jointly pre-trained on text, images, video (30T+ tokens, 2x Llama 3)
> MetaCLIP-based vision encoder, optimized for LLM integration
> Supports multi-image inputs (up to 8 tested, 48 pre-trained)
Long Context & iRoPE Architecture:
> 10M token support (Llama 4 Scout)
> Interleaved attention layers (no positional embeddings)
> Temperature-scaled attention for better length generalization
Training Efficiency:
> FP8 precision (390 TFLOPs/GPU on 32K GPUs for Behemoth)
> MetaP technique: Auto-tuning hyperparameters (learning rates, initialization)
Revamped Pipeline:
> Lightweight Supervised Fine-Tuning (SFT) → Online RL → Lightweight DPO
> Hard-prompt filtering (50%+ easy data removed) for better reasoning/coding
> Continuous Online RL: Adaptive filtering for medium/hard prompts
All model on Hugging Face - time to COOK!
Ragavan retweeted
BREAKING: Meta's Llama 4 Maverick just hit #2 overall - becoming the 4th org to break 1400+ on Arena!🔥
Highlights:
- #1 open model, surpassing DeepSeek
- Tied #1 in Hard Prompts, Coding, Math, Creative Writing
- Huge leap over Llama 3 405B: 1268 → 1417
- #5 under style control
Huge congrats to @AIatMeta — and another big win for open-source! 👏 More analysis below⬇️
Today is the start of a new era of natively multimodal AI innovation.
Today, we’re introducing the first Llama 4 models: Llama 4 Scout and Llama 4 Maverick — our most advanced models yet and the best in their class for multimodality.
Llama 4 Scout
• 17B-active-parameter model with 16 experts.
• Industry-leading context window of 10M tokens.
• Outperforms Gemma 3, Gemini 2.0 Flash-Lite and Mistral 3.1 across a broad range of widely accepted benchmarks.
Llama 4 Maverick
• 17B-active-parameter model with 128 experts.
• Best-in-class image grounding with the ability to align user prompts with relevant visual concepts and anchor model responses to regions in the image.
• Outperforms GPT-4o and Gemini 2.0 Flash across a broad range of widely accepted benchmarks.
• Achieves comparable results to DeepSeek v3 on reasoning and coding — at half the active parameters.
• Unparalleled performance-to-cost ratio with a chat version scoring ELO of 1417 on LMArena.
These models are our best yet thanks to distillation from Llama 4 Behemoth, our most powerful model yet. Llama 4 Behemoth is still in training and is currently seeing results that outperform GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on STEM-focused benchmarks. We’re excited to share more details about it even while it’s still in flight.
Read more about the first Llama 4 models, including training and benchmarks ➡️ go.fb.me/gmjohs
Download Llama 4 ➡️ go.fb.me/bwwhe9
Excited to share a research breakthrough from our team. It's fun, it's personal, it's customizable. Huge congrats to the team that worked hard to get to this milestone.
ai.meta.com/blog/movie-gen-m…
🎥 Today we’re premiering Meta Movie Gen: the most advanced media foundation models to-date.
Developed by AI research teams at Meta, Movie Gen delivers state-of-the-art results across a range of capabilities. We’re excited for the potential of this line of research to usher in entirely new possibilities for casual creators and creative professionals alike.
More details and examples of what Movie Gen can do ➡️ go.fb.me/kx1nqm
🛠️ Movie Gen models and capabilities
Movie Gen Video: 30B parameter transformer model that can generate high-quality and high-definition images and videos from a single text prompt.
Movie Gen Audio: A 13B parameter transformer model that can take a video input along with optional text prompts for controllability to generate high-fidelity audio synced to the video. It can generate ambient sound, instrumental background music and foley sound — delivering state-of-the-art results in audio quality, video-to-audio alignment and text-to-audio alignment.
Precise video editing: Using a generated or existing video and accompanying text instructions as an input it can perform localized edits such as adding, removing or replacing elements — or global changes like background or style changes.
Personalized videos: Using an image of a person and a text prompt, the model can generate a video with state-of-the-art results on character preservation and natural movement in video.
We’re continuing to work closely with creative professionals from across the field to integrate their feedback as we work towards a potential release. We look forward to sharing more on this work and the creative possibilities it will enable in the future.
Ragavan retweeted
(1/6): We are entering the Age of Global Resilience. And today with my partner and @generalcatalyst MD Paul Kwan, we’re publishing our thesis on what this is and why we’re all-in: generalcatalyst.com/perspect…
“Desktop” "Window" “File” “Program” "Download" "Copy/Paste" "Save"
"Website" "Tab" “Homepage” "Link" “Online” "Browse" "Search"
"App" "Homescreen" “Feed” “Notifications” "Swipe" "Share" "Message"
What are the AI-native frames that will define how consumers interact with intelligent software systems?
What are the AI-native nouns & verbs that will form the vocabulary of this next generation of products?
Are you building these today? We’d love to chat. Cc @generalcatalyst