4 ms·

OpenAI would naturally optimize for the tests published by Marcus as a critique of GPT-2, yet GPT-3 still fails physical reasoning spectacularly (the one test n

by abiro 6y ago

OpenAI would naturally optimize for the tests published by Marcus as a critique of GPT-2, yet GPT-3 still fails physical reasoning spectacularly (the one test needing casual reasoning the most). There are two broader points here: 1. The lack of independently verifiable evaluation metrics for these type of models should make everyone very skeptical. (Who can afford to retrain GPT-3 from scratch?) 2. I find it difficult to believe that smart people still insist that a model incapable of representing causal relationships can produce intelligent answers.

SpicyLemonZest 6y ago
(1) I certainly agree with. But Marcus doesn't claim skepticism about GPT-3s intelligence; he claims that his evaluation metrics definitively show it doesn't understand the text it outputs or know anything about the world. (2) is, I think, a misunderstanding. People who believe GPT-3 is producing intelligent answers generally believe it can represent causal relationships.
- abiro 6y ago
  Fair points. For the record, re 2: The GPT family of models (and all neural networks for that matter) can estimate P(X | Y), but have no way of computing whether X -> Y or X <- Y.
  - Veedrac 6y ago
    A computation can represent causality without being made of causality-neurons.
moyix 6y ago
> OpenAI would naturally optimize for the tests published by Marcus as a critique of GPT-2 It would be difficult for them to do so since Marcus's GPT2 critique came out after they collected the dataset for GPT3. Marcus's article: Jan 2020 GPT-3 dataset: "Table 2.2 shows the final mixture of datasets that we used in training. The CommonCrawl data was downloaded from 41 shards of monthly CommonCrawl covering 2016 to 2019"