The Platonic Representation Hypothesis
Huh, Minyoung, Cheung, Brian, Wang, Tongzhou, Isola, Phillip
–arXiv.org Artificial Intelligence
We argue that representations in AI models, particularly deep networks, are converging. First, we survey many examples of convergence in the literature: over time and across multiple domains, the ways by which different neural networks represent data are becoming more aligned. Next, we demonstrate convergence across data modalities: as vision models and language models get larger, they measure distance between datapoints in a more and more alike way. We hypothesize that this convergence is driving toward a shared statistical model of reality, akin to Plato's concept of an ideal reality. We term such a representation the platonic representation and discuss several possible selective pressures toward it. Finally, we discuss the implications of these trends, their limitations, and counterexamples to our analysis.
arXiv.org Artificial Intelligence
May-13-2024
- Country:
- Europe > Austria
- Vienna (0.14)
- North America > United States (0.28)
- Europe > Austria
- Genre:
- Research Report (1.00)
- Industry:
- Health & Medicine > Therapeutic Area (0.45)
- Technology: