|
09:00
|
Welcome & intro
-
Slava Voloshynovskiy
(UniGE)
()
|
|
09:30
|
Large Models in Low Precision: Power, Limits and Scaling Laws
-
Dan Alistarh
( ISTA, Austria)
()
|
|
10:40
|
--- Coffee break ☕️🫖🥐 ---
|
|
11:00
|
Short-term attention enables long term memorization & Stochastic activations
-
Hervé Jégou
(META)
()
|
|
|
09:00
|
--- ☕️ 🫖 🥐 ---
|
|
09:30
|
The Watermark Awakens
-
Pierre Fernandez
(Meta, France)
()
|
|
10:40
|
--- ☕️ 🫖 ---
|
|
11:00
|
The power of small language models
-
Pavlo Molchanov
(NVIDIA, USA)
()
|
|
|
10:45
|
Boat trip to Yvoire
()
|
|
|
09:00
|
--- ☕️🫖🥐 ---
|
|
09:30
|
Generative World Models for Sample-Efficient Reinforcement Learning
-
Eloi Alonso
Vincent Micheli
()
|
|
10:40
|
--- ☕️🫖 ---
|
|
11:00
|
Energy-Based Transformers and the Future of Scaling
-
Alexi Gladstone
(University of Illinois Urbana-Champaign/META, the USA)
()
|
|
|
09:00
|
--- ☕️🫖🥐 ---
|
|
09:30
|
DINOv3
-
Federico Baldassarre
(Meta, France)
()
|
|
10:30
|
How perception got solved (albeit unsatisfyingly)
-
Lucas Beyer
Xiaohua Zhaị
()
|
|
11:30
|
--- ☕️🫖 ---
|
|
11:45
|
Reasoning by thinking compositionally
-
Emmanuel Abbe
(EPFL / Apple, CH)
()
|
|
|
12:10
|
--- Lunch ---
|
|
13:10
|
AI through the Information-Theoretic lens
-
Slava Voloshynovskiy
(UniGE)
()
|
|
14:10
|
--- Coffee break☕️🫖 ---
|
|
14:30
|
TwinTURBO: Semi-Supervised Fine-Tuning of Foundation Models via Mutual Information Decompositions
-
Quetant Guillaume
(UNIGE)
()
|
|
15:00
|
Llama4 in 4 bits
-
Brian Pulfer
(University of Geneva)
()
|
|
15:30
|
--- Coffee break☕️🫖 ---
|
|
15:50
|
Diffusion LM and Compression
-
Pavlo Molchanov
(NVIDIA, USA)
()
|
|
17:00
|
--- Apero on Lake Board: discussions/networking/free time ---
|
|
18:00
|
Transfer to Social Event
(until 18:30)
()
|
|
18:30
|
--- Dinner ---
|
|
|
12:10
|
--- Lunch ---
|
|
13:40
|
Handcrafted vs. Learned: A Comparative Evaluation of Digital Watermarking Techniques
-
Vitaliy Kinakh
()
|
|
14:10
|
Attacking and Defending Vision Foundation Models
-
Yury Belousov
()
|
|
14:40
|
--- ☕️ 🫖 ---
|
|
15:00
|
TBD
-
Brian Pulfer
(University of Geneva)
()
|
|
15:30
|
--- ☕️ 🫖 ---
|
|
15:50
|
TBD
-
Seyed Mohsen Moosavi-Dezfooli
(Apple, CH)
()
|
|
17:00
|
--- ☕️ 🫖 ---
|
|
17:20
|
Round table - industrial perspective of model security and privacy
()
|
|
|
12:30
|
--- Lunch & Discussion ---
|
|
16:45
|
Boat trip -Return to Geneva
()
|
|
|
12:10
|
--- Lunch ---
|
|
13:40
|
Mini-Bang Detection and Interpretation: Benchmarking Reasoning on Emergent and Open-Ended Dynamics
-
Andrey Ustuzanin
(Constructor University, Germany)
()
|
|
14:30
|
--- ☕️🫖 ---
|
|
14:50
|
Improving mid-resource language performance in open LLMs
-
Hanna Yukhymenko
(Hugging Face, CH)
()
|
|
15:40
|
--- ☕️🫖 ---
|
|
16:00
|
TBD
-
Yury Belousov
()
|
|
16:30
|
Binary Diffusion Probabilistic Model
-
Vitaliy Kinakh
()
|
|
17:00
|
Discussion & Brainstorm & Free time
()
|
|
|
12:45
|
--- Lunch ---
|
|
14:00
|
World Models for Code
-
Taco Cohen
(Meta, the UK)
()
|
|
15:00
|
--- ☕️🫖 ---
|
|
15:20
|
The free transformer
-
François Fleuret
(University of Geneva /Meta, CH/France)
()
|
|
16:20
|
--- ☕️🫖 ---
|
|
16:40
|
Vision-language models: in-context learning and adaptation for radio-astronomy
-
Mariia Drozdova
(Universite de Geneve (CH))
()
|
|
17:10
|
Radio Astronomy in the Era of Large Models
-
Erica Lastufka
(University of Geneva, CH)
()
|
|
17:40
|
Summary/Conclusion/Outlook
-
Slava Voloshynovskiy
(UniGE)
()
|
|
18:05
|
--- Afterparty by subscription ---
|
|
