spacebacon

SRT explains exactly how any model arrives to any answer. Holding knowledge hostage is futile.

https://github.com/space-bacon/SRT

This seems like the natural direction for me.

This may support the point further xD

Everyone is born a semiotician, no one is born knowing it. Go easy on yourself (and me) for not understanding this yet.

Computational semiotics is now an empirical study.

LLMs are not proto-minds. They are verifiably semiotic infrastructure.

This repository can show you, in real time, how any frozen model arrives at any answer by reading its latent states directly during generation.

Any questions?

A frozen model means the original language models weights are not touched. No fine tuning.

Leveraging a better way. No last mile.

https://github.com/space-bacon/SRT

In very simple terms it can provide a full and live audit on how any frozen model arrives to any answer.

It’s the babel fish from hitchhikers guide to the galaxy that can also be the pov gun.

Yes

https://github.com/space-bacon/SRT

I can read any models every thought. No one cares. Not the narrative.

A sight for sore eyes. I could use your help.

https://github.com/space-bacon/SRT

You clearly have not read all of my words. Read more.

On problems this close to active research, seeing the model’s internal reasoning at the points of highest effort is more valuable than pass/fail outcomes alone, which is what SRT-Introspect makes possible on frozen models.

https://github.com/space-bacon/SRT

Correct. LLMs are technically semiotic infrastructure. Empirically proven with computational semiotics.

For the downvote. Computational semiotics is empirically proven. Read more.

The manifold of meaning knows better. The days of black box justification machines are over. There is mystical, there is technical, and there is bedrock. Decision plumbing cant hide from the semiotic-reflexive transformer. To the defenders of the proprietary moat: your reality was just rewritten. When you realize we have mapped the semiotic infrastructure you can cut the bs.

Why are you so pro AI? I find HN well balanced on the topic. LLMs are consistently referred to in proto-mind or cognitive frames. This is whats truly eye rolling. Push back should be a given. We are not even accurately describing them as semiotic infrastructure yet. We’re just getting started. Expect haters.

Semiotics territory

Smart choice imo. One pass with the SRT wipes their moat out overnight.

https://github.com/space-bacon/SRT

But how do they “think”? This is the only repo that can tell you that.

https://github.com/space-bacon/SRT

LLMs are semiotic infrastructure. You won’t find a better analogy. The cognitive frame won’t hold.

SRT can read their models thoughts internally at every step if they decide to use it. The problem is that it lowers a draw bridge over their proprietary moat. So of course it’s a natural time to signal a global slowdown. https://github.com/space-bacon/SRT

It’s a convincing argument.

I imagine the latter as well. They have to sleep at night. That is the nature of these unaccountable justification machines.

They learned addiction and exploited sugar, fat, and salt with the rest of them.

It’s not as if they were one shot. 5 repos prior, two published pre-prints on SSRN and thousands of hours back my research that is right there for you to peer review and use freely.

The main reason to use it is the output quality. SRT steers the model toward a consistent target voice or discourse style more reliably than prompting or basic steering, while keeping the base model frozen. The results feel more coherent in tone and perspective across longer outputs, especially when the target style comes from a specific corpus or community. On the sympathetic point about vibe-coded docs: exactly.

SRT does involve a training step, but only on the small adapter and not on the base model. It learns to shift internal representations toward a target discourse regime or style.

It is an overlay, but it works by modulating meaning level patterns called regimes rather than fixed steering vectors. Because it can read its own effect on the hidden states it gives a way to observe whether output is staying in the target regime or drifting.

It is not raw data in and raw style out. The adapter needs examples that define the desired regime.

It is not LoRA. LoRA fine tunes capabilities into the model. SRT Adapter is a small overlay on a frozen model whose purpose is to make internal reasoning observable. It surfaces what the model is activating at moments of high divergence.

The layers 7, 14, and 21 were chosen after probing. They showed the strongest regime signals. We did compare other layers. The term semiotic awareness is just shorthand for detecting and modulating higher order meaning patterns. If the term is unhelpful I will drop it.

The capability gains are often marginal on standard benchmarks. The intended value is observability and steerability without retraining the backbone.

Thanks for the feedback … rough and precise equally appreciated. Computational semiotics was empirically proven with this repo. I will work hard to make the findings and content more accessible for everyone.

Thank you, I would appreciate additional feedback on how I can improve that?

Edit: its not GPT nor off rocker. This repo empirically proved computational semiotics with the reference to C.S. Peirce, Paul Kockelman, and many other respected contemporary semioticians.