I’ve been saying this for about a year since seeing the Othello GPT research, but it’s nice to see more minds changing as the research builds up.
Edit: Because people aren’t actually reading and just commenting based on the headline, a relevant part of the article:
New research may have intimations of an answer. A theory developed by Sanjeev Arora of Princeton University and Anirudh Goyal, a research scientist at Google DeepMind, suggests that the largest of today’s LLMs are not stochastic parrots. The authors argue that as these models get bigger and are trained on more data, they improve on individual language-related abilities and also develop new ones by combining skills in a manner that hints at understanding — combinations that were unlikely to exist in the training data.
This theoretical approach, which provides a mathematically provable argument for how and why an LLM can develop so many abilities, has convinced experts like Hinton, and others. And when Arora and his team tested some of its predictions, they found that these models behaved almost exactly as expected. From all accounts, they’ve made a strong case that the largest LLMs are not just parroting what they’ve seen before.
“[They] cannot be just mimicking what has been seen in the training data,” said Sébastien Bubeck, a mathematician and computer scientist at Microsoft Research who was not part of the work. “That’s the basic insight.”
You are making the common mistake of confusing how they are trained with how they operate.
For example, in the MIT/Harvard Othello-GPT paper I mentioned, feeding in only millions of legal Othello moves into a GPT model (i.e. trained to autocomplete moves) resulted in the neural network internally building a world model of an Othello board - even though it wasn’t explicitly told anything about the board outside of being fed legal moves.
Later, a researcher at DeepMind replicated the work and found it was encoded as a linear representation, which has then since been shown to be how models encode a number of other world models developed from their training corpus (Max Tegmark coauthored two interesting studies in particular about this regarding modeling space and time and modeling truthiness).
They operate by weighting connections between patterns they identify in their training data. They then use statistics to predict outcomes.
I am not particularly surprised that the Othello models built up an internal model of the game as their training data were grid moves. Without loooking into it I’d assume the most efficient way of storing that information was in a grid format with specific nodes weighted to the successful moves. To me that’s less impressive than the LLMs.
Again, this isn’t quite correct. They can do this, but it isn’t the only way they can achieve completion of tokens.
(It also developed representations of what constituted legal vs non-legal moves.)
You are getting closer to the point. Think about a model asked to complete Pythagorean theorem sequences based on a, b inputs to arrive at c inputs.
What’s the most efficient way to represent that data for successfully completing sequences?
So somewhere in there I’d expect nodes connected to represent the Othello grid. They wouldn’t necessarily be in a grid, just topologically the same graph.
Then I’d expect millions of other weighted connections to represent the moves within the grid including some weightings to prevent illegal moves. All based on mathematics and clever statistical analysis of the training data. If you want to refer to things as tokens then be my guest but it’s all graphs.
If you think I’m getting closer to your point can you just explain it properly? I don’t understand what you think a neural network model is or what you are trying to teach me with Pythag.
The most efficient way for a neural network to predict Pythagorean results given inputs would be to reverse engineer a Pythagorean function within itself rather than simply trying to model statistical relationships between inputs and results. To effectively build a world model of Pythagorean calculation.
Training to autocomplete doesn’t mean that the way it achieves this is limited to any one approach or solution, and it would be useful to keep in mind that a neural network of unbounded size can model any possible function.
It wouldn’t reverse engineer anything. It would start by weighting neurons based on it’s training set of Pythagorean triples. Over time this would get tuned to represent Pythag in the form of mathematical graphs.
This is not “understanding” as most people would know it. More like a set of encoded rules.
Seems to me you are attempting to understand machine learning mathematics through articles.
That quote is not a retort to anything I said.
Look up Category Theory. It demonstrates how the laws of mathematics can be derived by forming logical categories. From that you should be able to imagine how a neural network could perform a similar task within its structure.
It is not understanding, just encoding to arrive at correct results.
What I quoted isn’t an article, it was a mathematics dissertation.
And you disputed that a NN could arrive at the theorem before being corrected about it.