←
UCL CSML277 posts
UCL CSML avatar

UCL CSML

@uclcsml

UCL CSML @ELLISforEurope Unit regrouping @GatsbyUCL, @uclcs and @stats_UCL. We organise the @JumpTrading CSML Seminar Series.

◷ Joined March 2012▧ 24 Photos and videos
UCL CSML avatar
UCL CSML@uclcsml·Oct 8

The talk will be based on the following work.

🔥Welcome 🔺Simplex Diffusion Models🔻! Standard discrete diffusion models suffer from "information collapse" because they discard uncertainty at intermediate steps by sampling categorical tokens. (1/5)

Post attachment
Valentin De Bortoli avatar
⟳ UCL CSML retweeted
Valentin De Bortoli@ValentinDeBort1·Sep 29

🔥Welcome 🔺Simplex Diffusion Models🔻! Standard discrete diffusion models suffer from "information collapse" because they discard uncertainty at intermediate steps by sampling categorical tokens. (1/5)

Post attachment
Arthur Gretton avatar
⟳ UCL CSML retweeted
Arthur Gretton@ArthurGretton·Sep 10

Applying to #UCL for a PhD in AI? Then be sure to apply for ELLIS as well! A great opportunity to collaborate with leading European AI labs.

🎓 Ready for a PhD in AI/ML? The #ELLISPhD Program connects top students with leading researchers across Europe. The application portal opens on 1 Oct. Register now to join the virtual info session on 30 Sep. All info at bit.ly/3TrH9CJ

Post attachment
ELLIS avatar
⟳ UCL CSML retweeted
ELLIS@ELLISforEurope·Sep 10

🎓 Ready for a PhD in AI/ML? The #ELLISPhD Program connects top students with leading researchers across Europe. The application portal opens on 1 Oct. Register now to join the virtual info session on 30 Sep. All info at bit.ly/3TrH9CJ

Post attachment
Antonin Schrab avatar
⟳ UCL CSML retweeted
Antonin Schrab@AntoninSchrab·Jul 23

New Aggregation Methods! -SingleBatch SB -TwoBatch TB -DataDriven SB -Sequential SB Applications -Multiple Hypothesis Testing -Conformal Prediction MMDAgg-SB HSICAgg-SB -Exact level -Higher power -Admit p-value -Minimax optimal Schrab* Shah Gretton Kim* arxiv.org/abs/2607.15823

Google Gemma avatar
⟳ UCL CSML retweeted
Google Gemma@googlegemma·Jun 10

Meet DiffusionGemma! An experimental open model that explores a fast approach to text generation, released under an Apache 2.0 license. Moving beyond sequential, token-by-token processes to generate entire blocks of text simultaneously. Here’s what’s new with DiffusionGemma: 👇

0:05
Krishna Balasubramanian avatar
⟳ UCL CSML retweeted
Krishna Balasubramanian@krizna_b·May 26

🚨 🧵 New paper: Finite-particle rates for drifting models. Drifting trains 1-step generators by moving model μ toward data ν with a conservative or non-conservative velocity field. Consider data on x-axis, model on the y-axis. These 2 fields yield 2 different motions. 1/7