Alex Bowe

Eng @NVIDIA | ex @GroqInc, Cruise | Compressed Data Structures PhD (BOSS de Bruijn Graph) | Photographer, gamer, learner, enjoyer of things

[0xkato @0xkato]

LLMs explained without all that yucky math stuff

https://t.co/cpgVukSbJx

[Noam Brown @polynoamial]

After AlphaGo, the skill of human Go players noticeably improved. I suspect we will see a similar pattern in math.

[Nick Lindquist @nick_lindquist]

Central Park is great, but it takes up a lot of space and isn’t utilized to its full potential.

That’s why I worked with McKinsey on a plan to make it a state of the art data center, complemented by rooftop parking and nuclear power.

We can still build beautiful things.

[priyanshu.sol @priyanshudotsol]

someone wrote a 680 page interactive book on cs algorithms

[Alex Bowe @alexbowe]

@weezerOSINT @_der_blAde_ I just started learning about BYOVD etc this week and find it fascinating - would love to know some of your favourite resources.

[ksa 🏴‍☠️ @kosa12m]

Best paper I've read so far this month:

All elementary functions (sin, cos, tan, exp, log, powers, roots, hyperbolic functions, π, e, and even basic arithmetic) can be generated from just one binary operator:
eml(x, y) = exp(x) − ln(y)
…plus the constant 1.

[shira @shiraeis]

Found a paper that suggests we may have spent years training agents to become hunters of proxy reward when the more basic thing intelligence craves is not a reward at all, but to not run out of viable futures.

The paper proposes that behavior is best understood as maximizing future action-state path occupancy, which collapses mathematically into a discounted entropy objective. The agent doesn’t necessarily want to GET something, but rather is trying to keep as many meaningful trajectories alive as possible.

The obvious objection is “so it just does random shit? fuck around and find out?”

No, this is where it gets pretty beautiful. The agent is variable when variation is cheap and becomes surgically goal-oriented the moment an absorbing state (death, starvation, falling over, etc) gets close enough to threaten its future path space.

Variability is the same drive as goal-directedness, just operating under different constraints.

The demos are kinda wild:

- A cartpole (classic move a cart to keep a pole from falling control task) that doesn’t merely balance but dances and swings through a huge range of angles and positions because why not? The whole point is occupying state space, and rigid balance is a voluntarily impoverished life.

- A prey-predator gridworld where the mouse PLAYS with the cat, teasing it and using both clockwise and counterclockwise routes around obstacles to lure it away from the food source before slipping in to eat, using both routes roughly equally. A reward-maximizing agent would collapse to one strategy and exploit it. Here, the agent keeps its behavioral repertoire

- A quadruped trained with Soft Actor-Critic and ZERO external reward that learns to walk, jump, spin, and stabilize, and then makes a beeline for food only when its internal energy drops low enough that starvation becomes a real threat

The thing that hit me hardest is the comparison to empowerment and free energy principle agents. Both collapse to near-deterministic policies with almost no behavioral variability. This paper’s agents find the highest-empowerment state and exploit it. FEP agents converge to classical reward maximizers.

As far as I’m aware, this is the only framework that produces agents you could describe as being “alive.”

The AI implication here is that we undertrain for behavioral repertoire. Most systems hit the benchmark by collapsing onto a narrow attractor basin of good-enough trajectories. They’re competent for sure, but brittle too, with one viable plan, executed until the world shifts and leaves them with nothing.

The thing I increasingly want from agents isn’t competence per se, but option-preserving competence.

I want agents with the ability to keep multiple viable plans alive and switch between them without catastrophe.

We’ve been so focused on teaching agents what to want that we never stopped to ask what happens if wanting isn’t the point, if the deepest drive isn’t necessarily toward anything, but away from the walls closing in.

paper: https://t.co/Kn3mllmmPK

[Mason Wang @masonwang025]

new post tomorrow!

[Alex Bowe @alexbowe]

Dumping ROMs using a microscope…

[azya @OneBitOnePixel]

McDonald's Chicken Nugget Tetris (2023) has now been dumped and emulated. This is a significant achievement for me, as the game's microcontroller is quite modern, and tackling something like this was beyond my reach until now.

[Alex Bowe @alexbowe]

@OneBitOnePixel @LowMax You dump ROMs by taking photos of them? That is the coolest thing I’ve heard all day.

[Alex Bowe @alexbowe]

@0thernet It’s a privilege to have annoyed users, especially ones who give you feedback

[Alex Bowe @alexbowe]

@0thernet The Siri one is so true. I’m textin this mf like it’s my personal assistant.

[Alex Bowe @alexbowe]

@0thernet @zocomputer Let's fucking zo

[Alex Bowe @alexbowe]

@OpenAI Please give it a Developer Console tool to automate reverse engineering REST APIs.

[Alex Bowe @alexbowe]

@evanjconrad You could probably just schedule tweet them at this point

[Alex Bowe @alexbowe]

@0thernet @zocomputer Zo far Zo good!

[Alex Bowe @alexbowe]

@crislenta @TheDeFiSalesman @zach_yadegari @naval What did he say?

[Alex Bowe @alexbowe]

@MisalignedModel @vitransformer @AmanGokrani Those are handy, but not all prompts and metadata are logged in the official logs. It’s better to use a proxy or monkey-patch the API calls (I’ve been using claude-trace by @badlogicgames)

[Alex Bowe @alexbowe]

@swyx Count me in! How do I sign up?

[Deedy @deedydas]

Huge computer science result:

A Tsinghua professor JUST discovered the fastest shortest path algorithm for graphs in 40yrs.

This improves on Turing award winner Tarjan’s O(m + nlogn) with Dijkstra’s, something every Computer Science student learns in college.