Alice Dauphin

PhD Student @TUGraz, going to @NeurIPSConf

[Guillaume Pourcel @GuillaumeAP]

🧠 Interested in local temporal learning in the brain? ✨ Come tomorrow to our spotlight talk at Neuromonster 🇮🇹! @thought_channel

@AliceDauphin4 and I will present recent follow-up work on Hamiltonian Echo Learning (Pourcel & Ernoult 2025) addressing key neuro questions...

❓ "Nice theory, but it only works for reversible, Hamiltonian dynamics, brains aren't." ❓"How does your theory relate to data? Your 'echo' phase looks a lot like hippocampal backward replays …" 👀 See you there 🥂

[Alice Dauphin @AliceDauphin4]

🛫 On my way to @NeurIPSConf in SD to present my poster @WiMLworkshop on Tue 2 Dec 6-9 pm PST!! I will present a novel dense Hopfield-like NN that can learn temporal dependencies with Hebbian rule using RHEL (check this great thread for a quick overview of what's RHEL👇)!

[Guillaume Pourcel @GuillaumeAP]

🧵Who said backprop was the only way to get gradients?
With Maxence Ernoult (@GoogleDeepMind), we investigated a hardware-friendly alternative that turns physical systems into self-learning machines: Recurrent Hamiltonian Echo Learning (RHEL)!
🔥RHEL @NeurIPS25 (oral)🔥

RHEL leverages the same physics for inference & learning. 🔁
Jump to the tweet [DEMO] for a simple interactive demo (no math 🙃, just intuitive mechanics ⚙️) of how RHEL turns a physical system into a self-learning machine 🕹️
📆 Meet us in person next week @EurIPSConf & @NeurIPSConf !!!
🇩🇰 Copenhagen Oral: Fri 5 Dec p.m CET
🇺🇲 San Diego
Oral: Fri 5 Dec 4:10 p.m. PST
Poster: Fri 5 Dec 4:30 p.m. PST
WiML Workshop Poster: Tue 2 Dec 6 pm PST (follow up neuro work for theorizing learning in the brain 🧠 presented by @AliceDauphin4)

[Guillaume Pourcel @GuillaumeAP]

@j_foerst To remove the hardware-software barrier, you need to tie inference and learning together. We found that a large class of physical systems (Hamiltonians) can simultaneously run inference AND compute their own cost gradients - no separate system for BPTT, no storing frwd activities

[Alice Dauphin @AliceDauphin4]

@FitzClaridge @Sover3ign1 @arjunkhemani @DavidDeutschOxf @NautilusMag @GuillaumeAP

[Alice Dauphin @AliceDauphin4]

@fchollet @GuillaumeAP

[Yoshua Bengio @Yoshua_Bengio]

Today, we are publishing the first-ever International AI Safety Report, backed by 30 countries and the OECD, UN, and EU.

It summarises the state of the science on AI capabilities and risks, and how to mitigate those risks. 🧵

Link to full Report: https://t.co/k9ggxL7i66

1/16

[Geoffrey Hinton @geoffreyhinton]

I think Elon Musk should be expelled from the British Royal Society. Not because he peddles conspiracy theories and makes Nazi salutes, but because of the huge damage he is doing to scientific institutions in the US. Now let's see if he really believes in free speech.

[Chris Olah @ch402]

Have feedback on Claude's behavior? Amanda leads the team that shapes Claude's personality.

[🇺🇦 Dzmitry Bahdanau @DBahdanau]

if you have no other joy in your life other than thinking of your importance as AI researcher

then GO FIND SOMETHING ELSE (e.g. sports, arts, friends, kids)

stay grounded and preserve your mental health

this may sound trivial, but given recent events, worth repeating

[Julien Pourcel @PourcelJulien]

1/ If you want to know how to generate diverse and challenging Programming Puzzles stop by our @NeurIPSConf poster #3010 (East Exhibit Hall A-C, 4:30-7:30 pm) @cedcolas @gaia_molinaro @pyoudeyer @lae_teo

[Alice Dauphin @AliceDauphin4]

@QuantaMagazine @StephenOrnes @GuillaumeAP

[Felix Hill @FelixHill84]

My solution

Just find a tiny problem that at least partially interests you

e.g. read papers (that have their code public) and think if there's any tiny change that could make the results different

Then write up the work as if it's a paper

Then (re)do your experiment

1/n

[Alice Dauphin @AliceDauphin4]

@GuillaumeAP @foresightinst @allisondman @RenPhil21 @MIT @AdamMarblestone @Convergent_FROs I am in

[galen @galenbrain]

I'm TAing a graduate course called Neural Computation this semester. It's a super fun class, taught by the brilliant Bruno Olshausen. I'm going to try to blog about the lectures! (link in bio)

[Alice Dauphin @AliceDauphin4]

@SachaIchbiah @GuillaumeAP

[Alice Dauphin @AliceDauphin4]

@GuillaumeAP @erikphoel @rob_lh looks amazing

[Alice Dauphin @AliceDauphin4]

@weijima01 @peiyuan9 @sheffdocfest @GuillaumeAP

[Alice Dauphin @AliceDauphin4]

@GuillaumeAP @Dexerto https://t.co/DrClMxRXon

[Alice Dauphin @AliceDauphin4]

@rmnth @LucCogZest @GuillaumeAP

[Alice Dauphin @AliceDauphin4]

@galenbrain **worship