Francesco Ruscio

PhD student in AI at #AITHYRA & @imperialcollege with @mmbronstein and @bose_joey | MSc Quantum at @ETH

[Francesco Ruscio @francescoshox]

Very happy to share the first paper of my PhD: Normalizing Flow Maps! (NFM)⚡️

We make likelihood evaluation in flow maps fast and tractable by replacing the DiT backbone with a normalizing flow🔄.

A big thanks to @tim_hitge, @mmbronstein, @bose_joey and the rest of the team!

[Joey Bose @bose_joey]

🚨 New Paper Alert, 🎤 Accepted for an oral presentation at the NeurIPS 2026 STODY workshop!

📣 Introducing: Normalizing Flow Maps (NFM): Combining few-step transport from Flow maps with tractable likelihoods from Normalizing Flows.

This is my first work with my first crop of new PhD students, @francescoshox and @tim_hitge, who genuinely knocked this project out of the park.

📄 Paper: https://t.co/LCFBg1QjLT
📝 Blog: https://t.co/8dWEYwUEOm
💻 Code: https://t.co/Oroppp6NBR

🧵
1/8

[Lucas José @lucasjvds]

The 3D codes that design fusion stellarators need exact solutions to check accuracy. Existing solutions covered at most 2 repeating segments, but real stellarators have 3 or more.

With Gemini 4 Carbon agents working together, we found solutions for any number N of segments.

[Philipp Nazari @philna00]

Due to (not so) popular demand, I decided to improve our main animation. Check out our paper! 👇

[Michael Bronstein @mmbronstein]

Stellar lineup at the AITHYRA Next generation AI for Biomedical Science workshop -- next Monday 2 Nov at the IMBA lecture hall in Vienna Biocenter. No registration is required.

[Romeo Passaro @rom_pas_]

Happy to share our new work: Distributional Functional Rewards (DFR)! 🎉

We improve diversity and fidelity in diffusion fine-tuning 🔧✨

Big thanks to my coauthors @JeongjaeLee0, Jong Chul Ye, @mmbronstein, @Yoshua_Bengio, @AlexanderTong7 and @danyalrehman17! 🙏 Thread below 👇

[Francesco Ruscio @francescoshox]

Check this one out!

[Jeongjae Lee @JeongjaeLee0]

🚨 New paper alert!
Despite recent advances in reward fine-tuning for diffusion and flow models, achieving high rewards remains challenging without sacrificing sample diversity or substantially drifting away from the pretrained distribution.
We address this with DFR: Distributional Functional Rewards, a framework that balances reward optimization with preserving desirable properties of the generated distribution! 🧵

[Joey Bose @bose_joey]

@CSProfKGD Finally have the talk up: https://t.co/RJMl1Hiip0

Enjoy! Feel free to ask any questions or feedback :)

[Zander Blasingame @zwblasingame]

Excited to announce the follow-up to our NeurIPS 2026 Oral: Improved Training of Strong Stochastic Flow Maps!

We improve the training of SSFMs by unlocking derivative-based losses!

Joint work with @antoniofrancaib @rmincam @ChristianKoke @Niklas_TR and @AlexanderTong7

[Nicolas Beltran @nicolasvelezb]

@francescoshox @giladturok Ah super neat! Thanks for sharing.

I'm excited to see if it would scale! It would totally change my view of the usefulness of likelihoods in flow based models.

[Francesco Ruscio @francescoshox]

@nicolasvelezb @giladturok You can check our recent paper: https://t.co/gMCdfEBs4n

It's still toyish (pixel space imagenet-64) but we use pixel space likelihoods

[Joey Bose @bose_joey]

🚨 New Paper Alert, 🎤 Accepted for an oral presentation at the NeurIPS 2026 STODY workshop!

📣 Introducing: Normalizing Flow Maps (NFM): Combining few-step transport from Flow maps with tractable likelihoods from Normalizing Flows.

This is my first work with my first crop of new PhD students, @francescoshox and @tim_hitge, who genuinely knocked this project out of the park.

📄 Paper: https://t.co/LCFBg1QjLT
📝 Blog: https://t.co/8dWEYwUEOm
💻 Code: https://t.co/Oroppp6NBR

🧵
1/8

[Francesco Ruscio @francescoshox]

@thoma_gu @bose_joey @tim_hitge @zwblasingame @nmboffi @AlexanderTong7 Oh no, we spent some time thinking about the acronym, how did we miss this? 🥶

[Nicholas Boffi @nmboffi]

more flow maps with exact likelihoods! was a ton of fun to work on this :)

[Zander Blasingame @zwblasingame]

Great work led by @francescoshox and @tim_hitge, glad I was able to help. Check out the details below:

[Francesco Ruscio @francescoshox]

@MauriceHanisch @tim_hitge @mmbronstein @bose_joey Thanks!

[Francesco Ruscio @francescoshox]

Very happy to share the first paper of my PhD: Normalizing Flow Maps! (NFM)⚡️

We make likelihood evaluation in flow maps fast and tractable by replacing the DiT backbone with a normalizing flow🔄.

A big thanks to @tim_hitge, @mmbronstein, @bose_joey and the rest of the team!

[Joey Bose @bose_joey]

🚨 New Paper Alert, 🎤 Accepted for an oral presentation at the NeurIPS 2026 STODY workshop!

📣 Introducing: Normalizing Flow Maps (NFM): Combining few-step transport from Flow maps with tractable likelihoods from Normalizing Flows.

This is my first work with my first crop of new PhD students, @francescoshox and @tim_hitge, who genuinely knocked this project out of the park.

📄 Paper: https://t.co/LCFBg1QjLT
📝 Blog: https://t.co/8dWEYwUEOm
💻 Code: https://t.co/Oroppp6NBR

🧵
1/8

[Francesco Ruscio @francescoshox]

@ArnaudDoucet1 @tim_hitge @mmbronstein @bose_joey Thank you very, there is still much to do!

[Niklas Rindtorff @Niklas_TR]

Introducing Langevin Flow Maps

TLDR: Neural Network Potentials that learn their own stochastic pathwise solution for faster simulation.

- Transferable across molecular systems.
- Transition dynamics are preserved.
- No trajectory data are required.

Work together with @sgmccallum, supervised by @AlexanderTong7, and @JamesFosterBath.

Arxiv: https://t.co/GGgujg56YH
Data: https://t.co/TeYvKN13o7
Code: https://t.co/HYiMqTI4jT

[Louis @louistichelman]

Can one foundation model reason across different graphs, features, relation types and prediction tasks?

Introducing 𝗪𝗮𝗻𝗱𝗲𝗿🚶

Our new preprint explores 𝗴𝗿𝗮𝗽𝗵 𝗳𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻 𝗺𝗼𝗱𝗲𝗹𝘀 through a shared prediction objective and a random-walk interface.

🧵 1/7

[Riccardo De Santi @r_de_santi]

We’re bringing natural-language control to any scientific generator.

We are very excited to introduce LLM-to-Flow Distillation (LFD), a method to adapt scientific generators to high-level natural-language discovery goals.

Starting from textual goals like “design peptides with promising anticancer activity” or “design small molecules binding to the D2 dopamine receptor”, LFD distills high-level goal-related knowledge from LLMs into scientific generators through iterative DPO on LLM-generated sub-goals.

🔥 LFD does not need task-specific proxy rewards, works with any frontier LLM via API-access, and outperforms state-of-the-art reward-guided methods such as FlowGRPO and DiffusionNFT on held-out computational evaluations. (1/6)

[David González-Martínez @_davidgonmar]

mHC and AttnRes demonstrate that richer depth mixing improves model performance at scale. Is there a unifying perspective? 🤔

Meet our Depth Memory framework: each layer reads from a set of streams to form its input, then writes back its output.

Linear Depth Memory (LDM) equips models with a matrix-valued memory.

[Alex Tong @AlexanderTong7]

We're recruiting PhD students at #AITHYRA in Vienna! Working on ML methods at the intersection of AI and biology.

Details & apply: https://t.co/UzneeVBnxl

Feel free to reach out if you have questions!