Dhruv Mahajan

Research Scientist at Meta AI. Interested in all things related to Computer Vision and Machine Learning. Father of 1. (Happily) Married

[Dhruv Mahajan @dkm2110]

Lets go Llama 4

[Ahmad Al-Dahle @Ahmad_Al_Dahle]

Introducing our first set of Llama 4 models!

We’ve been hard at work doing a complete re-design of the Llama series. I’m so excited to share it with the world today and mark another major milestone for the Llama herd as we release the *first* open source models in the Llama 4 collection 🦙. Here are some highlights:

📌 The Llama series have been re-designed to use state of the art mixture-of-experts (MoE) architecture and natively trained with multimodality. We’re dropping Llama 4 Scout & Llama 4 Maverick, and previewing Llama 4 Behemoth.

📌 Llama 4 Scout is highest performing small model with 17B activated parameters with 16 experts. It’s crazy fast, natively multimodal, and very smart. It achieves an industry leading 10M+ token context window and can also run on a single GPU!

📌 Llama 4 Maverick is the best multimodal model in its class, beating GPT-4o and Gemini 2.0 Flash across a broad range of widely reported benchmarks, while achieving comparable results to the new DeepSeek v3 on reasoning and coding – at less than half the active parameters. It offers a best-in-class performance to cost ratio with an experimental chat version scoring ELO of 1417 on LMArena. It can also run on a single host!

📌 Previewing Llama 4 Behemoth, our most powerful model yet and among the world’s smartest LLMs. Llama 4 Behemoth outperforms GPT4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on several STEM benchmarks. Llama 4 Behemoth is still training, and we’re excited to share more details about it even while it’s still in flight.

A big thanks to all of our launch partners (full list in blog) for helping us bring Llama 4 to developers everywhere including @huggingface, @togethercompute, @SnowflakeDB, @ollama, @databricks and many others👏 This is just the start, we have more models coming and the team is really cooking – look out for Llama 4 Reasoning 😉

A few weeks ago, we celebrated Llama being downloaded over 1 billion times. Llama 4 demonstrates our long-term commitment to open source AI, the entire open source AI community, and our unwavering belief that open systems will produce the best small, mid-size and soon frontier models. Llama would be nothing without the global open source AI community & we are so ready to begin this next chapter with you. 🦙

Read more about the release here: https://t.co/7mbK3uggjO, and try it in our products today.

[Ahmad Al-Dahle @Ahmad_Al_Dahle]

Introducing our first set of Llama 4 models!

We’ve been hard at work doing a complete re-design of the Llama series. I’m so excited to share it with the world today and mark another major milestone for the Llama herd as we release the *first* open source models in the Llama 4 collection 🦙. Here are some highlights:

📌 The Llama series have been re-designed to use state of the art mixture-of-experts (MoE) architecture and natively trained with multimodality. We’re dropping Llama 4 Scout & Llama 4 Maverick, and previewing Llama 4 Behemoth.

📌 Llama 4 Scout is highest performing small model with 17B activated parameters with 16 experts. It’s crazy fast, natively multimodal, and very smart. It achieves an industry leading 10M+ token context window and can also run on a single GPU!

📌 Llama 4 Maverick is the best multimodal model in its class, beating GPT-4o and Gemini 2.0 Flash across a broad range of widely reported benchmarks, while achieving comparable results to the new DeepSeek v3 on reasoning and coding – at less than half the active parameters. It offers a best-in-class performance to cost ratio with an experimental chat version scoring ELO of 1417 on LMArena. It can also run on a single host!

📌 Previewing Llama 4 Behemoth, our most powerful model yet and among the world’s smartest LLMs. Llama 4 Behemoth outperforms GPT4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on several STEM benchmarks. Llama 4 Behemoth is still training, and we’re excited to share more details about it even while it’s still in flight.

A big thanks to all of our launch partners (full list in blog) for helping us bring Llama 4 to developers everywhere including @huggingface, @togethercompute, @SnowflakeDB, @ollama, @databricks and many others👏 This is just the start, we have more models coming and the team is really cooking – look out for Llama 4 Reasoning 😉

A few weeks ago, we celebrated Llama being downloaded over 1 billion times. Llama 4 demonstrates our long-term commitment to open source AI, the entire open source AI community, and our unwavering belief that open systems will produce the best small, mid-size and soon frontier models. Llama would be nothing without the global open source AI community & we are so ready to begin this next chapter with you. 🦙

Read more about the release here: https://t.co/7mbK3uggjO, and try it in our products today.

[Dhruv Mahajan @dkm2110]

@ylecun @ylecun fortunately pineapple on a pizza can be unanimously characterized as grave human error rather than AI making a blunder.

[Yann LeCun @ylecun]

@yudapearl @GaryMarcus @TiernanRayTech 1. I respectfully suggest that you form opinions about my work from my own writings, not interviews, e.g. https://t.co/7ZgRtLIQWY
2. I do not understand why you think that ML/DL conflicts with causal inference. I see ML/DL as our best candidate to learn causal models.

[AI at Meta @AIatMeta]

Come join us at the XAI4CV workshop on Monday June 20th at #CVPR22 to learn about the latest research in explainable and transparent models. We have a fantastic set of invited speakers, 30+ posters, and a great panel discussion, details: https://t.co/VVf7ZxyoWX

[AI at Meta @AIatMeta]

Generalized Additive Models (GAMs) are fully interpretable ML models, unlike DNNs, but it's hard to make them efficient. We're sharing research on scaling GAMs to real world tasks w/o sacrificing accuracy or interpretability. https://t.co/9a9zaP4kdN https://t.co/1IxxVceeu2 [1/4]

[AI at Meta @AIatMeta]

Meta AI is calling researchers and developers to submit their papers to be highlighted at our workshop at @CVPR. Our workshop will touch on explainable #AI for computer vision. Learn more: https://t.co/VVf7ZxyoWX

[Dhruv Mahajan @dkm2110]

@karpathy In my opinion, overall number of data points seen (not batch) with throughput (data points processed per second) as the counter metric are the right metrics

[Devi Parikh @deviparikh]

XAI4CV workshop at @CVPR 2022! If you have any relevant work in explainable AI for computer vision (new work, or recently accepted at the main CVPR 2022 conference), consider submitting!

[Deepti Ghadiyaram @deeptigp]

We are excited to organize a workshop on Explainable AI for Computer Vision @ #CVPR2022! Our call for papers is open and we are actively sourcing submissions. More details here: https://t.co/B7PYqF6plW #cvpr