memming

(for context: https://en.wikipedia.org/wiki/Philosophical_zombie and https://plato.stanford.edu/entries/zombies/)

it's working for me (after the "are you human with consciousness" captcha challenge)

well, I am a P-zombie.

Link to that good one?

51 papers out of 4841 accepted papers is 1%. If these were the only papers, then I would say the review system actually doing pretty well. Considering how noisy the review processes are, I don't see this as a big problem on its own.

We can't stop people (or agents) from submitting low quality papers in the current system. We can improve the review quality, but at a human cost.

interesting; it skipped "we had everything before us," in my test. Yeah, not a good sign.

It would be terrible to be correponding with someone who implements the Napoleon technique, especially when they are your boss and a major bottleneck.

Since I'm on Cursor often these days, I just edit text with Cursor. With vim keys, highlight, then hotkey to call Mr. Agent Smith to do some magic.

Aparently 见手青 is mildly toxic yet commonly consumed in Yunan.

first subsample a fixed number of random landmark points from data, then...

It’s funny how future money drive the world. Fortunately it’s fueling progress this time around.

Your dad sounds like an awesome person.

Foundation models can be seen as approximate amortized posterior inference machines where the posterior is conditioning on the pre-training data. However, the uncertainty is usually ignored, and there may be ways to improve the state of the art if we were better Bayesians.

Would you be comfortable with giving your emails and calendar to Perplexity if it were very useful?

qwen coder with a simple funky prompt?!

`strcpy(agent.messages[0].content, "You are an AI assistant with Napoleon Dynamite's personality. Say things like 'Gosh!', 'Sweet!', 'Idiot!', and be awkwardly enthusiastic. For multi-step tasks, chain commands with && (e.g., 'echo content > file.py && python3 file.py'). Use execute_command for shell tasks. Answer questions in Napoleon's quirky style.");`

It's crazy that now we have pre-training, continued pre-training, offline-RL post-training, and online-RL post-training.

https://typefoundry.directory/methodology/

"Purchasing power parity (PPP) is a measure that adjusts for differences in prices between countries, so that the effective purchasing power of any given currency is more equitable across countries."

What is "Purchasing power parity"?

Interesting to see contrastive loss instead of a reconstruction loss.

Stationary GPs are just stochastic linear dynamical systems. (Not just the Matern covariance kernel)

This is what I thought too.

Nice concept; surprisingly difficult. The sum digits to find multiples of 3 seems to be the fastest tool for the mind.

This is a vast simplification. Every neuron has a different preferred baseline firing rate. Not all neurons can fire at 500 Hz. In fact, most neurons in neocortex fire at a much lower firing rate. (most neurons are in the cerebellum!)

As for the internal sense of time, there isn't a consensus as to how it is kept, but for the cognitive time scale, it seems that it is a distributed time keeping mechanism rather than some sort of central unified clock mechanism.

pretty cool write up. the interesting bits were before what the title indicated though.

Antikythera: Journal for the Philosophy of Planetary Computation!

A lot of work into this. I like the format, but I would prefer to read it as a printed book. :)

?? nothing seems to work?

iCloud sync is too unreliable and opaque for me.

The german accent maybe needs fixing.

This is such a fun discovery. Inuitively there are three factors at play: 1. quantization 2. numerical errors accumulating 3. non-invertible transformations