Harnessing the Universal Geometry of Embeddings · HackerPapers discussions

Harnessing the Universal Geometry of Embeddings — HN submission

Sun, 06 Sep 2026 20:31:20 GMT

Harnessing the Universal Geometry of Embeddings

122 points; 46 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Harnessing the Universal Geometry of Embeddings — mentioned by genxy

Mon, 17 Aug 2026 06:47:42 GMT

Is this the paper? Harnessing the Universal Geometry of Embeddings https://arxiv.org/abs/2505.12540

Paper and discussion history. A mention is not an endorsement.

Harnessing the Universal Geometry of Embeddings — mentioned by ainch

Tue, 07 Jul 2026 20:34:51 GMT

Thank you for expanding! I come from more of an ML background so still learning a lot on these topics: Agreed that the neocortex uses fewer layers because of looping - I also suspect it's partly because neurons are more complex than the neurons used in ANNs, so in principle they should be capable of more sophisticated computation in a single forward pass (especially considering that handling multiple neurotransmitters could mean superposed functions). The point about there being one correct way to model something does seem backed up by the Platonic Representation hypothesis…

Paper and discussion history. A mention is not an endorsement.

Harnessing the Universal Geometry of Embeddings — mentioned by observationist

Thu, 19 Mar 2026 15:50:13 GMT

https://arxiv.org/abs/2505.12540 https://arxiv.org/abs/2405.07987 These, other papers, and the lottery ticket phenomenon; what it boils down to is that any neural network like system which encodes some common mapping of a phenomenon in the context of the world - not necessarily a world model, but some "real-world thing" - will tend to map to a limited number of permutations of some archetypal representation, which will resemble other mappings of the same thing. The lottery ticket phenomenon is a bit like the birthday paradox; there will be some number of structures in a large, random…

Paper and discussion history. A mention is not an endorsement.

Harnessing the Universal Geometry of Embeddings — mentioned by mnky9800n

Fri, 30 Jan 2026 11:29:34 GMT

If you can access a models emebeddings then it is possible to retrieve what it knows using a model you have trained https://arxiv.org/html/2505.12540v2

Paper and discussion history. A mention is not an endorsement.

Harnessing the Universal Geometry of Embeddings — mentioned by AIorNot

Tue, 09 Dec 2025 01:51:16 GMT

Interesting - I wonder if this ties into the Platonic Space Hypothesis recently being championed by computational biologist Mike Levin E.g https://youtu.be/Qp0rCU49lMs?si=UXbSBD3Xxpy9e3uY https://thoughtforms.life/symposium-on-the-platonic-space/ e.g see this paper on Universal Embeddings https://arxiv.org/html/2505.12540v2 "The Platonic Representation Hypothesis [17] conjectures that all image models of sufficient size have the same latent representation. We propose a stronger, constructive version of this hypothesis for text models: the universal latent structure of text representations can…

Paper and discussion history. A mention is not an endorsement.

Harnessing the Universal Geometry of Embeddings — HN submission

Wed, 21 May 2025 18:15:25 GMT

Harnessing the Universal Geometry of Embeddings

123 points; 40 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Harnessing the Universal Geometry of Embeddings — HN submission

Tue, 20 May 2025 21:39:43 GMT

Text Embeddings are All Alike

5 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.