Visualizing Properties of Encoders

[robf]

@garry.morrison

a while back I generated so called “word classes” just using the similarity of pre and post superpositions. And indeed, it worked really quite well at finding words that could be substituted for each other. But is there a sensible way to improve even further on this idea?

You can do like a transformer, and “learn” which contexts to attend to with “attention”. 10 billion dollars in investment from Microsoft to OpenAI says clearly it works quite well.

Or you can leave all the information in a network, and let the runtime context select what it wants to attend to.

To expand on that second solution…

I think the bigger problem than identifying context to generalize on, and even bigger than the entanglement of generalizations which Coecke etc. have focused on, is the fact that these generalizations actually act like expansions of the data, and there appears to be no limit to them. So any fixed set will always be incomplete, and the only way to completely capture them, is to find the ones relevant to a given situation at run time. I went into this a lot in the “How Your Brain Organizes Information” thread. This post might be a good summary of (the first part of) that:

"How Your Brain Organizes Information" video

Let’s use the word “factorize” as you use it: OK. So we’ve got “Language … can be decomposed into … a mathematical expression.” What is that mathematical expression? I say it can’t. I say language (“natural” human language, to distinguish it from computer or artificial languages) cannot be completely described any more compactly than a full body of its usage (that’s what distinguishes natural language from computer language or mathematical languages, and what has prevented us from building…

I gave a lot of examples, from the history of linguistics, maths, and philosophy, to support the idea that meaning can’t be completely abstracted. That in fact “learning”, acts like an expansion of the data. But I thought the data that… @cezar_t presented was also good support - that doing more training over transformers, acts much like just using more data:

@robf

If parameters were a simplification of the data you might expect that the number of meaningful parameters would decrease as the data size decreased. If you’re simplifying something it seems reasonable to assume it results in something smaller. But this seems to be indicating “simplifying” is much the same as just increasing the data size. With this result you might argue that calculating more parameters “expanded” structure to a degree roughly comparable (4x) to the degree to which the data size was reduced (1/4.) Calculating more parameters wasn’t simplifying more, it was expanding in roughly the same way as adding data.

I’m hypothesizing the models will continue to get better as the number of parameters are increased, even if data size is limited. This seems to me to be consistent with that

So that’s the first option. “Learn” the context to “attend” to. And capture the entangled, quantum, quality of generalizations, by having a black box, where no-one is sure what the structure is.

But I say it runs into the bigger problem, that generalizations expand. They are not only entangled as Coecke etc see, but they get forever bigger. The current solution to that seems to be to just try to have it as big as possible. (Which also means that only the biggest entities can get involved at all, and the little guy is reduced to begging, or paying, to get access to big company APIs.)

The real flaw with that is that no matter how big you make it, it will never be as “big” as human performance, and you’re forever chasing an asymptote:

@robf

I see what Tesla is doing, heaps of money thrown at artificial situation generators, to generate beforehand every conceivable thing that might happen, so every conceivable thing can be “learned” before it happens.

Or language transformers, only novel by interpolation over stuff that’s been said before.

So much money and effort being wasted.

Elon Musk talks to the F

The second option, one to deal with this expansion, the one I’m now focused on, Is to leave all the information in a network, leave it “embodied” in a set of data, essentially, and let the runtime context select what it wants to attend to. As I said earlier in this thread:

@robf

I now think that generalization on the embedding was a mistake. Even though my vector product was generating new structure, I had a fixed embedding frozen in the definition of my vectors.

I now think that is a problem inherent to vectors. The only way to keep all the information is to keep it in a network. Hence what I’ve been talking about trying to do in recent threads here, projecting out “embeddings” directly from a network.

In the network formulation the “cross-product” becomes just a recombination of elements based on an “embedding” which varies at run-time.

But I’ve talked a lot about that in my other threads.

This explanation to @JarvisGoBrr might be a good summary:

The coding of longer sequences in HTM SDRs

For the record, I posse exactly the same question You said it yourself Jarvis: So we’re agreed word embeddings work. The only problem is how to extract them from a network of sequences: I start from the position that such sequences are related in the network by sharing (multiple) beginning and end points. That’s the shape of the network. If that’s the shape of the network, why shouldn’t we be able to tune the network to reveal that shape. It’s there. We just have to figure out how to p…

Together with this:

The coding of longer sequences in HTM SDRs

Nice. But what kind of similarity are the neurons clustered by? I’ll reference this reply to @cezar_t which I thought was a clear explanation too. And it nests two or three references to the structure I’m looking for in other threads too. These things get lost in the length of the thread(s): I expect the words of the input to cluster into (relative) lines in a similar way. So one way to think about this is to say I want to replicate the Brazilian paper, but do it for the clusters formed b…