Not at all - Transformer was invented by a bunch of former Google employees (while at Google), primarily Jakob Uszkoreit and Noam Shazeer. Of course as with anything it builds on what had gone before, but it's really quite a novel architecture.
The scientific impact of the transformer paper is large, but in my opinion the novelty is vastly overstated. The primary novelty is adapting the (already existing) dot-product attention mechanism to be multi-headed. And frankly, the single-head -> multi-head evolution wasn't particularly novel -- it's the same trick the computer vision community applied to convolutions 5 years earlier, yielding the widely-adopted grouped convolution. The lasting contribution of the Transformer paper is really just ordering the existing architectural primitives (attention layers, feedforward layers, normalization, residuals) in a nice, reusable block. In my opinion, the most impactful contributions in the lineage of modern attention-based LLMs are the introduction of dot-product attention (Bahdanau et al, 2015) and the first attention-based sequence-to-sequence model (Graves, 2013). Both of these are from academic labs.
As a side note, a similar phenomenon occurred with the Adam optimizer, where the ratio of public/scientific attribution to novelty is disproportionately large (the Adam optimizer is very minor modification of the RMSProp + momentum optimization algorithm presented in the same Graves, 2013 paper mentioned above)
I think the most novel part of it, and where a lot of the power comes from, is in the key based attention, which then operationally gives rise to the emergence of induction heads (whereby pair of adjacent layers coordinate to provide a powerful context lookup and copy mechanism).
The reusable/stackable block is of course a key part of the design since the key insight was that language is as much hierarchical as sequential, and can therefore be processed in parallel (not in sequence) with a hierarchical stack of layers that each use the key-based lookup mechanism to access other tokens whether based on position or not.
In any case, if you look at the seq2seq architectures than preceded it, it's hard to claim that the Transformer is really based-on/evolved-from any of them (especially prevailing recurrent approaches), notwithstanding that it obviously leveraged the concept of attention.
I find the developmental history of the Transformer interesting, and wish more had been documented about it. It seems from interview with Uszkoreit that the idea of parallel language processing based on an hierarchical design using self-attention was his, but that he was personally unable to realize this idea in a way that beat other contemporary approaches. Noam Shazeer was the one who then took the idea and realized it in the the form that would eventually become the Transformer, but it seems there was some degree of throw the kitchen sink at it and then a later ablation process to minimize the design. What would be interesting to know would be an honest assessment of how much of the final design was inspiration and how much experimentation. It's hard to imagine that Shazeer anticipated the emergence of induction heads when this model was trained at sufficient scale, so the architecture does seem to at least partly be an a accidental discovery, and more than the next generation seq2seq model that it seems to have been conceived as.
At least, the Transformer didn't. The abstract idea of a language model goes way back though within the field of linguistics, and people were building simplistic "N-gram" models before ever using neural nets, then using other types of neural net such as LSTMs and CNNs(!) before Google invented the Transformer (primarily with the goal of fully utilizing the parallelism available from GPUs - which couldn't be done with a recurrent model like LSTM).
Not at all - Transformer was invented by a bunch of former Google employees (while at Google), primarily Jakob Uszkoreit and Noam Shazeer. Of course as with anything it builds on what had gone before, but it's really quite a novel architecture.
The scientific impact of the transformer paper is large, but in my opinion the novelty is vastly overstated. The primary novelty is adapting the (already existing) dot-product attention mechanism to be multi-headed. And frankly, the single-head -> multi-head evolution wasn't particularly novel -- it's the same trick the computer vision community applied to convolutions 5 years earlier, yielding the widely-adopted grouped convolution. The lasting contribution of the Transformer paper is really just ordering the existing architectural primitives (attention layers, feedforward layers, normalization, residuals) in a nice, reusable block. In my opinion, the most impactful contributions in the lineage of modern attention-based LLMs are the introduction of dot-product attention (Bahdanau et al, 2015) and the first attention-based sequence-to-sequence model (Graves, 2013). Both of these are from academic labs.
As a side note, a similar phenomenon occurred with the Adam optimizer, where the ratio of public/scientific attribution to novelty is disproportionately large (the Adam optimizer is very minor modification of the RMSProp + momentum optimization algorithm presented in the same Graves, 2013 paper mentioned above)
I think the most novel part of it, and where a lot of the power comes from, is in the key based attention, which then operationally gives rise to the emergence of induction heads (whereby pair of adjacent layers coordinate to provide a powerful context lookup and copy mechanism).
The reusable/stackable block is of course a key part of the design since the key insight was that language is as much hierarchical as sequential, and can therefore be processed in parallel (not in sequence) with a hierarchical stack of layers that each use the key-based lookup mechanism to access other tokens whether based on position or not.
In any case, if you look at the seq2seq architectures than preceded it, it's hard to claim that the Transformer is really based-on/evolved-from any of them (especially prevailing recurrent approaches), notwithstanding that it obviously leveraged the concept of attention.
I find the developmental history of the Transformer interesting, and wish more had been documented about it. It seems from interview with Uszkoreit that the idea of parallel language processing based on an hierarchical design using self-attention was his, but that he was personally unable to realize this idea in a way that beat other contemporary approaches. Noam Shazeer was the one who then took the idea and realized it in the the form that would eventually become the Transformer, but it seems there was some degree of throw the kitchen sink at it and then a later ablation process to minimize the design. What would be interesting to know would be an honest assessment of how much of the final design was inspiration and how much experimentation. It's hard to imagine that Shazeer anticipated the emergence of induction heads when this model was trained at sufficient scale, so the architecture does seem to at least partly be an a accidental discovery, and more than the next generation seq2seq model that it seems to have been conceived as.
1 reply →
This makes no sense. A thing's roots don't change, either it did start there or it didn't.
It didn't.
At least, the Transformer didn't. The abstract idea of a language model goes way back though within the field of linguistics, and people were building simplistic "N-gram" models before ever using neural nets, then using other types of neural net such as LSTMs and CNNs(!) before Google invented the Transformer (primarily with the goal of fully utilizing the parallelism available from GPUs - which couldn't be done with a recurrent model like LSTM).