Shikai Li

Meta SuperIntelligence Labs.
Keep coding. Keep Learning.

[Hakm @hakmgpt]

To clarify things: I may have been unfair to the model.

Usually, when a new model drops, everyone tests it to its full potential. My previous prompt was also weak, so I improved it and tested Muse Wonder Max again.

I know I might get some hate for admitting this, but I was unfair to the model. It’s still not on Opus’s level, in my opinion, but I didn’t test it properly the first time.

[Mark Zuckerberg @finkd]

Introducing Muse Wonder, our new family of frontier models with state of the art performance for personal and professional agentic work.

[Claire Zhou @claireszhou]

Latest @SemiAnalysis_ article comparing coding agent subscription plans show that @Muse Code's $50 plan delivers 51.3x the value per dollar, topping the API value leaderboard.

The $50 plan gives a comparable amount of token value as both the Codex $200 and Claude $200 plans. For good coding performance (give max reasoning effort a spin!) + affordable and fast inference, it's hard to argue anything beats Muse Code right now.

To users hitting limits on this tier, we hear you on wanting a higher subscription option and will be sharing more on this soon!

Download Muse Code at https://t.co/0gA0vrHMFF

Full SemiAnalysis post: https://t.co/KYYYOc0fRh

[Shikai Li @levendlee]

@plumberbutt97 @alexeheath @finkd So much insider information 👀

[Zhiqing Sun @EdwardSun0909]

Which model is benchmaxxing?

If a model is good across different measurements, maybe it’s just good

[Alexandr Wang @alexandr_wang]

1/ we just publicly released Muse Spark 1.3 max!

we see significantly stronger coding and agentic performance on muse spark 1.3 max, so would strongly recommend trying it out even if you've already tried muse spark 1.3 high or muse spark 1.3 xhigh.

[Mark Zuckerberg @finkd]

Muse Spark 1.3 is rolling out today with frontier performance almost too cheap to meter. This is the biggest jump we've made so far on coding and agentic work. Try it in Muse Code and our API.

Next up 🍉 and Muse Spark open weights releases coming soon.

[Alexandr Wang @alexandr_wang]

good iteration for muse spark 1.2

onto bigger and better things 🍉

[Alexandr Wang @alexandr_wang]

muse code in beta is live. first coding agent from msl, built on muse spark 1.2.

install: curl -fsS https://t.co/TFXsXfJKW8 | bash

here's what you should know:

[Alexandr Wang @alexandr_wang]

1/ today we're releasing muse spark, the first model from MSL. nine months ago we rebuilt our ai stack from scratch. new infrastructure, new architecture, new data pipelines. muse spark is the result of that work, and now it powers meta ai. 🧵

[PyTorch @PyTorch]

Mixture-of-Experts (MoE) is a popular #LLM architecture that reduces computation by activating fewer parameters per token. But it brings memory, communication, and control challenges.

💡We introduce MetaShuffling, enabling efficient Llama 4 model inference in production.🔗 Read our latest blog to learn more: https://t.co/bxp2kPE5n1

#PyTorch #OpenSourceAI

[Shikai Li @levendlee]

Time to deep dive again.

[Rihanna @rihanna]

what happened yesterday in Atlanta was brutal, tragic & is certainly not an isolated incident by any means. AAPI hate has been rampantly perpetuated & it’s disgusting! I’m heartbroken for the Asian community & my heart is with the loved ones of those we lost. The hate must stop.

[Kamala Harris @KamalaHarris]

A man followed a Chinese-American doctor “spewing a profanity-laced racist tirade as she walked to the subway. ‘Why are you Chinese people killing everyone?’”

There is no place for hatred in our country and these racist attacks must be condemned.
https://t.co/ymyKH6JgR6

[Shikai Li @levendlee]

@ttq1008 哈哈，短暂的三个月

[Shikai Li @levendlee]

Quarantine provides a good opportunity to study and improve yourself.

[Barack Obama @BarackObama]

Watch this. It shows why we should all do the right thing and stay home to the fullest extent possible. All of us can help slow the spread of the virus, protecting the elderly, the vulnerable, and each other. https://t.co/FgffQrMVB7

[News from Google @NewsFromGoogle]

“We are fully aligned and continue to work with the US Government to contain the spread of COVID-19, inform citizens, and protect the health of our communities. (1/6)

[Shikai Li @levendlee]

咔嚓！