This couldn't be more perfectly timed .. I have an Unreal Engine game with both VT100 terminals (for running coding agents) and Z80 emulators, and a serial bridge that allows coding agents to program the CP/M machines:
Those "coincidences" in Connections are really no coincidence at all, but path dependence. Breakthrough advance A is impossible or useless without prerequisites B and C and economic conditions D, but once B and C and D are in place, A becomes obvious next step.
If one would train an actual secret (e.g. a passphrase) into such a model, that a user would need to guess by asking the right questions. Could this secret be easily reverse engineered / inferred by having access to models weights - or would it be safe to assume that one could only get to the secret by asking the right questions?
I don’t know, but your question reminds me of this paper which seems to address it on a lower level: https://arxiv.org/abs/2204.06974
“Planting Undetectable Backdoors in Machine Learning Models”
“ … On the surface, such a backdoored classifier behaves normally, but in reality, the learner maintains a mechanism for changing the classification of any input, with only a slight perturbation. Importantly, without the appropriate "backdoor key", the mechanism is hidden and cannot be detected by any computationally-bounded observer. We demonstrate two frameworks for planting undetectable backdoors, with incomparable guarantees. …”
Luckily I have a very large amount of MSX computers, zx, amstrad cpc etc and even one multiprocessor z80 cp/m machine for the real power. Wonder how gnarly this is going to perform with bankswitching though. Probably not good.
Awesome. I've just designed and built my own z80 computer, though right now it has 32kb ROM and 32kb RAM. This will definitely change on the next revision so I'll be sure to try it out.
Pretty cool! I wish free-input RPGs of old had fuzzy matchers. They worked by exact keyword matching and it was awkward. I think the last game of that kind (where you could input arbitrary text when talking to NPCs) was probably Wizardry 8 (2001).
There are two things happening here. A really small LLM mechanism which is useful for thinking about how the big ones work, and a reference to the well known phenomenon, commonly dismissively referred to as a "trick", in which humans want to believe. We work hard to account for what our conversational partner says. Language in use is a collective cultural construct. By this view the real question is how and why we humans understand an utterance in a particular way. Eliza, Parry, and the Chomsky bot at http://chomskybot.com work on this principle. Just sayin'.
It depends on the model, but from my experiments (quantizing one layer of a model to 2-bit and then training the model with that layer in 2-bit to fix the damage) the first layer is the most sensitive, and yes, the last layer is also sensitive too. The middle layers take the best to quantization.
Different components of a layer also have a different sensitivity; e.g. the MLP downscale block damages the model the most when quantized, while quantizing the Q projection in self attention damages the model the least.
Nice - that will fit on a Gameboy cartridge, though bank switching might make it super terrible to run. Each bank is only 16k. You can have a bunch of them, but you can only access one bank at a time (well, technically two - bank 0 is IIRC always accessible).
It's pretty obvious this is just a stress test for compressing and running LLMs. It doesn't have much practical use right now, but it shows us that IoT devices are gonna have built-in LLMs really soon. It's a huge leap in intelligence—kind of like the jump from apes to humans. That is seriously cool.
i'll echo that practicality only surfaces once it is apparent what can be done. yea this feels like running "DOOM on pregnancy test devices" type of moment
interesting, i am wondering how far can it go if we remove some of these limitations but try to solve some extremely specific problem like generating regex based on user input? i know small models(270M range) can do that but can it be done in say < 10MB range?
Generate an LLM that is designed to solve one extremely specific problem: answering the ultimate question of life, the universe, and everything.
Even with modern supercomputing the computation would be outpaced by the heat death of the universe, so token output must be limited to a single integer.
For future projects and/or for this project, there are many LLMs available more than good enough to generate that kind of synthetic data (20 Qs) with permissive terms of use. (So you don’t need to stress about breaking TOS / C&D etc)
All the 'Small' language models and the 'TinyML' scene in general tend to bottom out at a million parameters, hence I though 'micro' is more apt at ~150k params.
I love these thought experiments. Looking at the code size, it would have been possible for someone to come up with this back in the days, similar to the idea of a million monkeys on a typewriter eventually producing Shakespeare.
Speaking of - I remember my first digital camera (Fujitsu 1Mb resolution using SmartMedia)… it used so much power that you could take 20-30 photos and then needed to replace all 4 batteries lol
This couldn't be more perfectly timed .. I have an Unreal Engine game with both VT100 terminals (for running coding agents) and Z80 emulators, and a serial bridge that allows coding agents to program the CP/M machines:
https://i.imgur.com/6TRe1NE.png
Thank you for posting! It's unbelievable how someone sometimes just drops something that fits right into what you're doing. However bizarre it seems.
Oh dear, it seems we've... somehow been psychically linked...
I developed a browser-based CP/M emulator & IDE: https://lockboot.github.io/desktop/
I was going to post that instead, but wanted a 'cool demo' instead, and fell down the rabbit hole.
Connections: Alternative History of Technology by James Burke documents these "coincidences".
Those "coincidences" in Connections are really no coincidence at all, but path dependence. Breakthrough advance A is impossible or useless without prerequisites B and C and economic conditions D, but once B and C and D are in place, A becomes obvious next step.
In before AI companies buy up all the Z80s and raise the prices to new heights.
If one would train an actual secret (e.g. a passphrase) into such a model, that a user would need to guess by asking the right questions. Could this secret be easily reverse engineered / inferred by having access to models weights - or would it be safe to assume that one could only get to the secret by asking the right questions?
I don’t know, but your question reminds me of this paper which seems to address it on a lower level: https://arxiv.org/abs/2204.06974
“Planting Undetectable Backdoors in Machine Learning Models”
“ … On the surface, such a backdoored classifier behaves normally, but in reality, the learner maintains a mechanism for changing the classification of any input, with only a slight perturbation. Importantly, without the appropriate "backdoor key", the mechanism is hidden and cannot be detected by any computationally-bounded observer. We demonstrate two frameworks for planting undetectable backdoors, with incomparable guarantees. …”
> this secret be easily reverse engineered / inferred by having access to models weights
It could with a network this small. More generally this falls under "interpretability."
Luckily I have a very large amount of MSX computers, zx, amstrad cpc etc and even one multiprocessor z80 cp/m machine for the real power. Wonder how gnarly this is going to perform with bankswitching though. Probably not good.
Awesome. I've just designed and built my own z80 computer, though right now it has 32kb ROM and 32kb RAM. This will definitely change on the next revision so I'll be sure to try it out.
RAM is very expensive right now.
We're talking kilobytes, not gigabytes. And it isn't DDR5 either.
Pretty cool! I wish free-input RPGs of old had fuzzy matchers. They worked by exact keyword matching and it was awkward. I think the last game of that kind (where you could input arbitrary text when talking to NPCs) was probably Wizardry 8 (2001).
There are two things happening here. A really small LLM mechanism which is useful for thinking about how the big ones work, and a reference to the well known phenomenon, commonly dismissively referred to as a "trick", in which humans want to believe. We work hard to account for what our conversational partner says. Language in use is a collective cultural construct. By this view the real question is how and why we humans understand an utterance in a particular way. Eliza, Parry, and the Chomsky bot at http://chomskybot.com work on this principle. Just sayin'.
As far as I know, the last layer is very quantization-sensitive, and is typically not quantized, or quantized lightly.
Have you experimented with having it less quantized, and evaluated the quality drop?
Regardless, very cool project.
(Not OP)
It depends on the model, but from my experiments (quantizing one layer of a model to 2-bit and then training the model with that layer in 2-bit to fix the damage) the first layer is the most sensitive, and yes, the last layer is also sensitive too. The middle layers take the best to quantization.
Different components of a layer also have a different sensitivity; e.g. the MLP downscale block damages the model the most when quantized, while quantizing the Q projection in self attention damages the model the least.
Nice - that will fit on a Gameboy cartridge, though bank switching might make it super terrible to run. Each bank is only 16k. You can have a bunch of them, but you can only access one bank at a time (well, technically two - bank 0 is IIRC always accessible).
It's pretty obvious this is just a stress test for compressing and running LLMs. It doesn't have much practical use right now, but it shows us that IoT devices are gonna have built-in LLMs really soon. It's a huge leap in intelligence—kind of like the jump from apes to humans. That is seriously cool.
i'll echo that practicality only surfaces once it is apparent what can be done. yea this feels like running "DOOM on pregnancy test devices" type of moment
This is super cool. Would love to see a Z80 simulator set up with these examples to play with!
interesting, i am wondering how far can it go if we remove some of these limitations but try to solve some extremely specific problem like generating regex based on user input? i know small models(270M range) can do that but can it be done in say < 10MB range?
Generate an LLM that is designed to solve one extremely specific problem: answering the ultimate question of life, the universe, and everything.
Even with modern supercomputing the computation would be outpaced by the heat death of the universe, so token output must be limited to a single integer.
Eliza's granddaughter.
did you measure token/s?
Meanwhile, Eliza was ported to BASIC and was run on many home computers in the 80s.
For future projects and/or for this project, there are many LLMs available more than good enough to generate that kind of synthetic data (20 Qs) with permissive terms of use. (So you don’t need to stress about breaking TOS / C&D etc)
An LLM in a .com file? Haha made my day
SLM
All the 'Small' language models and the 'TinyML' scene in general tend to bottom out at a million parameters, hence I though 'micro' is more apt at ~150k params.
Imagine, this working on a Gameboy, in those days. Would've sounded like magic
I don’t think this could beat an ELIZA-style bot in how magical it feels, given the extreme terseness of its replies.
I love these thought experiments. Looking at the code size, it would have been possible for someone to come up with this back in the days, similar to the idea of a million monkeys on a typewriter eventually producing Shakespeare.
And would have lasted 3 minutes.
Speaking of - I remember my first digital camera (Fujitsu 1Mb resolution using SmartMedia)… it used so much power that you could take 20-30 photos and then needed to replace all 4 batteries lol