@_arohan_ @bilaltwovec It means they are using something else?
rohan anil@_arohan_
Claude Fable let me implement Muon, Shampoo and K-FAC, what does this mean?
12:24 AM · Jun 10, 2026 · 573 Views
Simo Ryu argues Claude Fable's ability to generate complex ML optimizers like Muon points to alternative internal optimization
The model successfully generated advanced Shampoo and K-FAC optimization code
2223133320.6K
Sentiment
Positive reactions highlight interest in sharing sparse autoencoder research involving Muon and Adam after Claude Fable assisted with Muon, Shampoo, and K-FAC implementations, whereas critics claim the work sabotages research.
Pos
50.0%
Neg
50.0%
2 comments with sentiment.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS1.4KLIKES6REPLIES1
Dan Roy@roydanroy
@_arohan_ Haha.
rohan anil@_arohan_
Claude Fable let me implement Muon, Shampoo and K-FAC, what does this mean?
3hViews 1.4KLikes 6Bookmarks 0
BOOKMARKS1
K@k0x3k
@_arohan_ Does that mean it’s not frontier 😒🥲?
11hViews 651Bookmarks 1
RETWEETS3
rohan anil@_arohan_
Claude Fable let me implement Muon, Shampoo and K-FAC, what does this mean?
12hViews 19.3KLikes 222Bookmarks 33
ueaj@_ueaj
@_arohan_
12hViews 867Likes 3
Adam Mainz@MainzOnX
@_arohan_ Oof your kernel probably could have just been a PyTorch aten op call
11hViews 683Likes 2
Olcan@olcan
@_arohan_ maybe their detector failed, maybe those things are not at the frontier, maybe it silently sabotaged your implementations
12hViews 706Likes 1
Guru Chahal🇺🇸@guruchahal
@_arohan_ Sin gu lari T
12hViews 393Likes 1
Frosty40@FrostForger
actively sabotaging research. I cant man... your muon shamploo super soaker sodapop works when a major company declares war on its own customers in the form of subtle sabotage. also we know, your optimizer is the best. anyone arguing is still at whatever lab they chose, and not at CA, so yours will be the best. but we got some serious problems out here on the frontlines. I fucking switched careers(lol i dont have a career in ai), and study everyday to know about my future, and be a father who can lead his kids safely into an unknown future. Fuck the money man, this is so much bigger than money now.
12hViews 1.2K
*maxwell twelftree@mxwellllll
@_arohan_ i am writing up something on using sparse auto encoders within Muon and Adam, would be keen for you to take a look :)
11hViews 962
immortal@immortaldip
@_arohan_ > LLM research question detected > PEFT gets loaded > PEFT decide to tell Joke Muon, Shampoo, and K-FAC walk into a training run.
Adam looks over and says, “Great… the second-order optimizer support group is here to tell me I’m just momentum with marketing.”
12hViews 704
Reyaa@snr_boost
@_arohan_ It let you do particle physics, chemistry and advanced math?
12hViews 444
Santosh Mohan@theycallmeMohan
@_arohan_ Doesn’t it silently degrade performance in this case instead of flagging it explicitly to the user?
4hViews 89Likes 1
Kaizhao Liang@KyleLiang5
@_arohan_ Try AdamW 😂 see if it asks you “who is Adam”
4hViews 71Likes 1
Mayz@lunan_ai
@_arohan_ that list reads like someone just dumped their gradient descent notebook
which one should i be most worried about understanding?
11hViews 187
ar0cket1@ar0cket1
@_arohan_ its modified without notice, so no surprise
10hViews 147
Gauri Tripathi@Gauri_the_great
@_arohan_ something is fishy
7hViews 131
T Ay.@ayedtay
@_arohan_
3hViews 9
Amine Ben Khalifa@aminert
@_arohan_ Not frontier haha
3hViews 4
Blissy@BlissyOnX
@_arohan_ ngl that reads like a chaos scroll through optimization research
u forking the llm mafia or what
12hViews 2
Invincible@InvincibleEdge
@_arohan_ means u let a vibe cod ER build logic brick piles until architecture calc bc like 101 way back oh what they know early era 47 is there|they learning heavy optimization via formal gradient align align
12h
Load more posts