I don't want to dismiss this outright but I'm skimming this paper and pretty skeptical of something that's from a single guy that doesn't appear peer reviewed, spends most of its time talking about actual biology, comes up with a "RELU6" (RELU but minimum value 6), and then pushes detailed review to a future paper.
Figure 3 B in "Cooperation is All You Need" shows the same score curves as the top left of Figure 6 in "Beyond Attention," so it must be basically the same implementation. Yet that earlier paper is only cited once, in the Acknowledgements section. As far as I can tell, the only mathematical change in this paper is capping the ReLU at 6. But it also adds a bunch of grandiose verbiage ("triadic modulation loops", "awake thought.")
The author is clearly a crackpot. Maybe he wasn't a crackpot when he still managed to publish in peer-reviewed journals, but cognitive decline over time is not exactly unheard of.
Sat here vibe coding a pure assembly kernel for arm64, APL layer with conceptual memory layout. On my bed, eating a bag of chips, jobless since Jan. Everything but the Ket are mine
Who knows, but drop the word "vibe" and this is basically the startup culture 15 years ago, so ¯\_(ツ)_/¯.
Well, okay, for better historical accuracy, replace APL with API, and the kernel for arm64 thing with Ruby on Rails on a new Macbook, but the point still stands.
Probably as long as non-anthropomorphic idiocy can.
No opinion on this submission, but a more general point. I'm not the one to jump into anthropomorphizing computers, but last year or two of LLM and adjacent research is a constant stream of papers and experiments that totally surprise everyone who refuse to even entertain comparisons between LLMs and people, while being entirely expected and completely not surprising to those who do.
I did a chat with Gemini about the paper, and tldr is...
* They introduce a loop at the beginning between Q, K, and V vectors (theoretically representing "question", "clues" and "hypothesis" of thinking)
* This loop contains a non linearity (ReLU)
* The loop is used to "pre select" relevant info
* They then feed that into a light weight attention mechanism.
They claim OOM faster learning, and robustness acro domains. There's enough detail to probably do your own PuTorch implementation, though they haven't released code. The paper has been accepted into AMLDS2025. So peer reviewed.
At first blush, this sounds really exciting and if results hold up and are replicated, it could be huge.
https://arxiv.org/pdf/2305.10449
And this paper in an IEEE journal:
https://arxiv.org/pdf/2211.01950
The author is clearly a crackpot. Maybe he wasn't a crackpot when he still managed to publish in peer-reviewed journals, but cognitive decline over time is not exactly unheard of.
Ketamine is all you need
Well, okay, for better historical accuracy, replace APL with API, and the kernel for arm64 thing with Ruby on Rails on a new Macbook, but the point still stands.
Also
> Competing interests: AA has a provisional patent application for the algorithm used in this paper.
No opinion on this submission, but a more general point. I'm not the one to jump into anthropomorphizing computers, but last year or two of LLM and adjacent research is a constant stream of papers and experiments that totally surprise everyone who refuse to even entertain comparisons between LLMs and people, while being entirely expected and completely not surprising to those who do.
They claim OOM faster learning, and robustness acro domains. There's enough detail to probably do your own PuTorch implementation, though they haven't released code. The paper has been accepted into AMLDS2025. So peer reviewed.
At first blush, this sounds really exciting and if results hold up and are replicated, it could be huge.