Charlie Landau

Philosophy B.S. learning about ML

[Charlie Landau @adamwslave]

@sasuke___420 @dileeplearning 🤣🤣

[Charlie Landau @adamwslave]

@the_art_of_Li @Hesamation Grok is being pretty sycophantic. This is word salad

[Charlie Landau @adamwslave]

@sasuke___420 @dileeplearning Dileep is*

[Charlie Landau @adamwslave]

@sasuke___420 @dileeplearning Dileep’s talking about the LZ algorithm, not gzip. Idealized LZ provably converges to the entropy rate with no constant gap; which is exactly what makes it asymptotically optimal and cat not (cat’s rate never approaches H)

[Charlie Landau @adamwslave]

@l0tkas @captgouda24 That’s literally exactly what it does (disclaimer that I’ve taken linear algebra and econometrics courses before AI could handle them)

[Xingyi Yang @yxy2168]

SAE interventions can be unreliable. 🧠🔒

We show that even when features are clamped, bad behaviors can still return through alternative residual-space directions. 🧩↩️

Feature control ≠ behavior control. 🚨

Paper: https://t.co/6soSFXKI5N
Page+Code: https://t.co/tvg1lEky07

[Charlie Landau @adamwslave]

@the_art_of_Li @Hesamation Better prediction = intelligence = compression (more or less)

[Charlie Landau @adamwslave]

@the_art_of_Li @Hesamation In a more general sense, minimizing KL-divergence is the act of approaching the “true” distribution of the document/collection of documents, which is directly equivalent to better compression of those documents. LLM loss minimization is essentially better compression.

[Charlie Landau @adamwslave]

@pl_zeng I’d imagine this would be the case for any identity element of an operation. Do you guys intend on testing multiplication with a holdout on 1?

[Charlie Landau @adamwslave]

@lthlnkso @uncledoomer That’s a scary number you’ve got there

[Charlie Landau @adamwslave]

@decloaking @ns123abc @InferenceKD @theo He gets $ from engagement, it pays to be/act stupid

[Charlie Landau @adamwslave]

@ns123abc @InferenceKD @theo Youre right, you didn’t force or bait. You straight up lied

[Charlie Landau @adamwslave]

Bonus: First reply is author engaging with Holocaust denial. Clear he only cares about miscarriages of justice when the perpetrators are brown.

[Charlie Landau @adamwslave]

“Believe all women” applies to individuals making claims. Doubting the 250k number isn’t doubting women who’ve claimed to be SAed, it’s doubting the extrapolation (which is clearly smudged to anyone with half a brain).

[Steve's Law 🇬🇧 @MaximusChud]

"There's no way 250k girls were raped by Pakistanis over decades" says the side that believes 1/4 women have been raped.

It's "believe all women" until the women say they perputrator was brown.

Exhibit a: (wait until the end)

[Charlie Landau @adamwslave]

@RighttoTryGuy @AndyMasley They want a global slowdown. It doesn’t work if countries like China can continue researching while we’ve stifled development domestically.

[Charlie Landau @adamwslave]

@oprydai @NewPlace2Frown We were talking about this a couple days ago 🤣

[pulsatingGenius @Neuralwarlock]

https://t.co/3oOkzxFcV7

[Charlie Landau @adamwslave]

Opsec nuts with 0 trust in third party ID collection need to be studied

[Ahmad @TheAhmadOsman]

I immediately called that the next up from Anthropic will be ID Verification to use a closed AI model

Your freedoms will be taken one thing at a time if Local and Opensource AI doesn't win

[Marcin Szewczyk @wodny]

@IntCyberDigest Also remember there are oil deposits under the LinkedIn HQ.