3 ms·

This sounds like you consider what has been called "feature engineering" in older ML research as being fundamental to modern ML successes. But the trajectory ha

This sounds like you consider what has been called "feature engineering" in older ML research as being fundamental to modern ML successes. But the trajectory has been the opposite direction, away from feature engineering, and instead towards letting the transformer architecture learn as much as possible, including interesting features. Most research on deep net architecture, is about how to train it effectively on huge stacks of GPUs and TPUs. That's how/why transformers were invented.

mjburgess 4y ago
weights are just compressions of the training set. The reason hardware improvements have improved ML is just because we're compressing the training set to larger sizes. Going eg., from 100TB training to 300BB weights. The process of training is more-or-less just "building a search index", and prediction is little more than "looking up examples in the search index and combining".
- visarga 4y ago
  Reminds me of Pedro Domingos's paper - Every Model Learned by Gradient Descent Is Approximately a Kernel Machine https://arxiv.org/abs/2012.00152 https://arxiv.org/abs/2012.00152
  - YorkshireSeason 4y ago
    Modern DL uses stochastic gradient descent, which Domingos's paper doesn't deal with, in fact he leaves the generalisation as an open problem. Be that as it may, universality in computing models is a well known issue. I could also say "Every Model Learned by Gradient Descent Is an x86 processor + a lot of RAM", or ""Every Model Learned by Gradient Descent Is a 2 Layer Neural Net" (as 2 layer nets are universal approximators), but this is not terribly illuminating.
- YorkshireSeason 4y ago
  I don't understand those objections. Yes, one of the key insights of modern DL-based ML is that size matters, neural networks start producing interesting generalisation only once you have large parameter spaces. Indeed, one might wonder if size is all that matters? I don't see what is wrong with search indices, every computer is a gigantic search index, because main memory is a map from an index (the memory address) to 2^n bit integers, the value stored at the index. So anything that can be computed at all, can be computed by "a search index". The important thing that, prima facie, distinguish NNs from eg SQL databases, is the use real numbers as latent space: because real numbers are a metric space, so you have a notion of distance which allows you do use efficient local search like gradient descent. Moreover, when you ask this 'mere' search index (NNs) about something that it was not trained on, it can (and probably does) return something that is close in this metric space to something it learned. That did not work at all in the 1990s. Today it works spectacularly well in some domains. So an interesting question now is, to use your language: is human intelligence anything more than "building a search index" and "looking up examples in the search index and combining"? So this is a variation of Searle's Chinese room, but unlike him in 1980, we have working Chinese rooms!
  - mjburgess 4y ago
    It is important to note that the search space is just the training data. That space does not span enough dimensions to actually solve the problems in question. For example, we act in the world so as to set the state of possible causal variables -- and thereby can distinguis causes. Above, graph algorithms operate on a discrete graph space and hence on the graph. GPT here isn't doing that. Seeing NNs as distance-based searches in a compressed 'training space' demystifies what solutions can be found. Here the text space defined by the training set does not permit a graph traversal, so eg., the NN 'uses' edges that don't exist We are not searches across such spaces, which are radically impoverished and not parametrised by time or action
    - YorkshireSeason 4y ago
      It is not the case that "the search space is just the training data", the search space in NNs are some variants of real multivariate functions. The training data can be all manner of things, in LLMs, they are text in natural language. Graphs or natural language text are mapped from a discrete space to a continuous metric space. It is true that LLMs cannot currently solve complex mathematics problems reliably. Whether that's an intrinsic shortcoming of transformers, or not is an interesting open question, that is being investigated as we speak. A lot of well-known DL research teams bet on transformers beating humans in mathematics in the next few years, see e.g. https://imo-grand-challenge.github.io/
- nl 4y ago
  > The process of training is more-or-less just "building a search index", and prediction is little more than "looking up examples in the search index and combining". It's really not. You are compressing the information in that dataset. That means the model has to learn characteristics of the information in the data that represent it. This is one of the reasons why some argue that compression is a form of intelligence[1] It's a very different mechanism to a search index or database even if in some case it appears to work similarly. [1] http://www.hutter1.net/ai/ http://www.hutter1.net/ai/

mjburgess 4y ago

weights are just compressions of the training set. The reason hardware improvements have improved ML is just because we're compressing the training set to larger sizes. Going eg., from 100TB training to 300BB weights. The process of training is more-or-less just "building a search index", and prediction is little more than "looking up examples in the search index and combining".

visarga 4y ago
Reminds me of Pedro Domingos's paper - Every Model Learned by Gradient Descent Is Approximately a Kernel Machine https://arxiv.org/abs/2012.00152 https://arxiv.org/abs/2012.00152
- YorkshireSeason 4y ago
  Modern DL uses stochastic gradient descent, which Domingos's paper doesn't deal with, in fact he leaves the generalisation as an open problem. Be that as it may, universality in computing models is a well known issue. I could also say "Every Model Learned by Gradient Descent Is an x86 processor + a lot of RAM", or ""Every Model Learned by Gradient Descent Is a 2 Layer Neural Net" (as 2 layer nets are universal approximators), but this is not terribly illuminating.
YorkshireSeason 4y ago
I don't understand those objections. Yes, one of the key insights of modern DL-based ML is that size matters, neural networks start producing interesting generalisation only once you have large parameter spaces. Indeed, one might wonder if size is all that matters? I don't see what is wrong with search indices, every computer is a gigantic search index, because main memory is a map from an index (the memory address) to 2^n bit integers, the value stored at the index. So anything that can be computed at all, can be computed by "a search index". The important thing that, prima facie, distinguish NNs from eg SQL databases, is the use real numbers as latent space: because real numbers are a metric space, so you have a notion of distance which allows you do use efficient local search like gradient descent. Moreover, when you ask this 'mere' search index (NNs) about something that it was not trained on, it can (and probably does) return something that is close in this metric space to something it learned. That did not work at all in the 1990s. Today it works spectacularly well in some domains. So an interesting question now is, to use your language: is human intelligence anything more than "building a search index" and "looking up examples in the search index and combining"? So this is a variation of Searle's Chinese room, but unlike him in 1980, we have working Chinese rooms!
- mjburgess 4y ago
  It is important to note that the search space is just the training data. That space does not span enough dimensions to actually solve the problems in question. For example, we act in the world so as to set the state of possible causal variables -- and thereby can distinguis causes. Above, graph algorithms operate on a discrete graph space and hence on the graph. GPT here isn't doing that. Seeing NNs as distance-based searches in a compressed 'training space' demystifies what solutions can be found. Here the text space defined by the training set does not permit a graph traversal, so eg., the NN 'uses' edges that don't exist We are not searches across such spaces, which are radically impoverished and not parametrised by time or action
  - YorkshireSeason 4y ago
    It is not the case that "the search space is just the training data", the search space in NNs are some variants of real multivariate functions. The training data can be all manner of things, in LLMs, they are text in natural language. Graphs or natural language text are mapped from a discrete space to a continuous metric space. It is true that LLMs cannot currently solve complex mathematics problems reliably. Whether that's an intrinsic shortcoming of transformers, or not is an interesting open question, that is being investigated as we speak. A lot of well-known DL research teams bet on transformers beating humans in mathematics in the next few years, see e.g. https://imo-grand-challenge.github.io/
nl 4y ago
> The process of training is more-or-less just "building a search index", and prediction is little more than "looking up examples in the search index and combining". It's really not. You are compressing the information in that dataset. That means the model has to learn characteristics of the information in the data that represent it. This is one of the reasons why some argue that compression is a form of intelligence[1] It's a very different mechanism to a search index or database even if in some case it appears to work similarly. [1] http://www.hutter1.net/ai/ http://www.hutter1.net/ai/