Erlangen AI Hub

[levent @__alpoge__]

hello there the jacobian conjecture is false thanx to my close friend akhil for asking about it and my other close friend fable for working during the world cup final

((1+xy)^3 z + y^2 (1+xy) (4+3xy), y + 3 x (1+xy)^2 z + 3 x y^2 (4+3xy), 2 x - 3 x^2 y - x^3 z): \C^3\to \C^3, has jacobian determinant -2, and sends (0, 0, -1/4), (1, -3/2, 13/2), and (-1, 3/2, 13/2) to (-1/4, 0, 0)

[Erlangen AI Hub @ErlangenAIHub]

Mathematical Foundations of AI:
The Erlangen AI Hub Conference 2026
📍 Venue: Mathematical Institute, University of Oxford 📅 Date: 1–3 September 2026
��� Poster submissions are now open
🎟️ Early Bird tickets are available for a limited time
👉 Register below
@UniofOxford

[Erlangen AI Hub @ErlangenAIHub]

https://t.co/fOZky0wJSy

[Erlangen AI Hub @ErlangenAIHub]

📢"The Erlangen AI Hub is an asset to @UniofOxford , to its academic, commercial and institutional collaborators; and to the UK, which must succeed within a global, competitive, community." Great piece by Peter Grindrod, Professor at @OxUniMaths - 👇

[Ayhan Suleymanzade @ayhozade]

Can LLMs reason in superposition? We introduce MUX, a method that turns text CoT into latent continuous reasoning.

Instead of one-hot vectors as in CoT, the model now learns to predict weighted averages of several one-hot vectors, that we call multiplexed tokens. These multiplexed tokens can be designed to be lossless, so by predicting them one is essentially doing multi-token prediction (MTP) in superposition.

MUX is the best latent reasoning method across 32 math settings spanning 1-8B LLaMA base models, reducing CoT length by 3-6x. Furthermore, it is able to perform parallel search, harnessing a core strength of superposed reasoning.

In collaboration with @alperen_gozeten , @mmbronstein, @ismaililkanc, and @jw9730.

1/🧵

[Erlangen AI Hub @ErlangenAIHub]

https://t.co/bIxDhHmyel

[Erlangen AI Hub @ErlangenAIHub]

Huge congratulations to @wooldridgemike and Leslie Ann Goldberg on their election as Fellows of the @royalsociety 👏👏👏
See story below 👇

[Erlangen AI Hub @ErlangenAIHub]

https://t.co/vDgX5XSDRT

[Erlangen AI Hub @ErlangenAIHub]

As AI advances rapidly, our understanding struggles to keep pace. Mathematics plays a critical role in closing this gap.

Erlangen AI Hub founding director @mmbronstein will speak at our annual conference Mathematical Foundations of AI on September 1-3. See link below:

[Erlangen AI Hub @ErlangenAIHub]

Mathematical Foundations of AI:
The Erlangen AI Hub Conference 2026
📍 Venue: Mathematical Institute, University of Oxford 📅 Date: 1–3 September 2026
📌 Poster submissions are now open
🎟️ Early Bird tickets are available for a limited time
👉 Register below
@UniofOxford

[Erlangen AI Hub @ErlangenAIHub]

Some excellent ideas in this article.

[University of Oxford @UniofOxford]

‘Many of today’s most transformative technologies rest on mathematical ideas that once seemed abstract or esoteric.’

Why maths is pivotal for all AI technologies, and how future leadership will depend on investing in numerical disciplines.

Find out more⬇️

[Erlangen AI Hub @ErlangenAIHub]

... We will next explore how the foundations of AI are unlocking the next generation of AI systems.
📍 Venue: Mathematical Institute, University of Oxford
📅 Date: 1–3 September 2026
🎟️ Some Early Bird tickets still available
�� Register here: https://t.co/kcyqGpgP3L

[Erlangen AI Hub @ErlangenAIHub]

Stronger mathematical foundations lead to more reliable AI and a new community is forming around these ideas, centred on the work of the Erlangen AI Hub, with partners at 8 universities across the UK. After a very successful first annual conference ...(1/2) @mmbronstein