This raises an unavoidable question: why would a public defense proceed in private, even after the institution had been alerted to serious procedural and governance concerns?

UPDATE (30 October 2025):

The public livestream for this public defense was not activated. The event therefore took place in private, shielding it from the public scrutiny it is meant to invite.

🙏 Huge thanks to the fantastic team: @florianmahner.bsky.social, @lukasmut.bsky.social & @martinhebart.bsky.social — and all supporters at @mpicbs.bsky.social & @dondersinst.bsky.social!

#CognitiveNeuroscience #DeepLearning #NeuroAI

Dimensions underlying the representational alignment of deep neural networks with humans - Nature Machine Intelligence An interpretability framework that compares how humans and deep neural networks process images has been presented. Their findings reveal that, unlike humans, deep neural networks focus more on visual ...

🔗 Full text: doi.org/10.1038/s422...
🛠️ Code + data: github.com/florianmahne...

🥼 What we did
• Treated DNNs as humans in 20M odd-one-out trials
• Learned shared embeddings with interpretable dimensions (color, taxa, texture…)

🔍 Key finding
• Humans weight semantic cues first
• DNNs weight visual cues first

➡️ Why it matters
• Reveals a fundamental gap in human–AI alignment

🚨 Our new paper is out in @natmachintell.nature.com!
🤔 Deep neural nets (DNNs) often align with humans in performance and even representations — but do they "think" like us?
📄 “Dimensions underlying the representational alignment of deep neural networks with humans”

🚀 Hope this brings us one step closer to restoring meaningful vision to those who need it most.

#BrainComputerInterfaces #ComputableAwards #DigitalInnovation #NeuroAI #Neuroprosthetics #Neurotech #eLife

✨ Fun fact: Ours was the only award where both the jury and the public agreed! Such rare consensus made the recognition feel even more special.

🙏 Immensely grateful for the inspiring collaboration, and proud to see our work recognized on such a prestigious stage.

Towards biologically plausible phosphene simulation for the differentiable optimization of visual cortical prostheses The provided framework based on biological models and clinical literature can aid in the optimization of visual cortical prostheses through computational or behavioral simulation experiments, serving ...

🧠 Together, we introduced a biologically plausible phosphene simulation for the differentiable optimization of visual cortical prostheses. You can read more in our eLife paper: doi.org/10.7554/eLif...

👉 This award recognizes our groundbreaking work carried out by Maureen van der Grinten and Jaap de Ruyter van Steveninck, in close collaboration with NIN and UMH—especially Antonio Lozano—under the supervision of Yağmur Güçlütürk and myself.

📣 We won the "Dutch ICT Oscar"!

🏆 At the end of last year, our team was honored with the Computable Award in Digital Innovation 2024, “for visual prosthetics for the blind.”

Just hit 5K citations. Beyond thankful. To be continued. 🧠🤖🚀 #NeuralEncoding #NeuralDecoding #NeuroAI

Hello, World!

After years of broadcasting into the void, I'm retuning my antenna:

Professional signal → Bluesky & LinkedIn (linkedin.com/in/umguec/)
Personal noise → Instagram (instagram.com/umguec/)

Let the decentralized brain dump commence. 🧠💩