Abstract
AbstractModern large language models generate texts that are virtually indistinguishable from those written by humans and achieve near-human performance in comprehension and reasoning tests. Yet, their complexity makes it difficult to explain and predict their functioning. We examined a state-of-the-art language model (GPT-3) using lexical decision tasks widely used to study the structure of semantic memory in humans. The results of four analyses showed that GPT-3’s patterns of semantic activation are broadly similar to those observed in humans, showing significantly higher semantic activation in related (e.g., “lime–lemon”) word pairs than in other-related (e.g., “sour–lemon”) or unrelated (e.g., “tourist–lemon”) word pairs. However, there are also significant differences between GPT-3 and humans. GPT-3’s semantic activation is better predicted by similarity in words’ meaning (i.e., semantic similarity) rather than their co-occurrence in the language (i.e., associative similarity). This suggests that GPT-3’s semantic network is organized around word meaning rather than their co-occurrence in text.
Publisher
Springer Science and Business Media LLC
Reference35 articles.
1. Brown, T. B. et al. (2020). Language models are few-shot learners. arXiv http://arxiv.org/abs/2005.14165 (2020).
2. Van Noorden, R. How language-generation AIs could transform science. Nature 605(7908), 21–21. https://doi.org/10.1038/d41586-022-01191-3 (2022).
3. DeepL. (n.d.). DeepL SE. https://www.DeepL.com/translator
4. Bender, E. M., Gebru, T., McMillan-Major, A. & Shmitchell, S. On the dangers of stochastic parrots: Can language models be too big?. In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency 610–623. https://doi.org/10.1145/3442188.3445922 (2021).
5. Lake, B. M. & Murphy, G. L. Word meaning in minds and machines. Psychol. Rev. https://doi.org/10.1037/rev0000297 (2021).
Cited by
14 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献