←
Guy Van den Broeck84 posts
Guy Van den Broeck avatar

Guy Van den Broeck

@guyvdb

Professor of Computer Science at UCLA @UCLAComSci; working on Artificial Intelligence

◷ Joined April 2008▧ 2 Photos and videos
Metin Alp avatar
⟳ Guy Van den Broeck retweeted
Metin Alp@alpmetin1·Nov 12

Make sure to check out our paper!

📄 CoRL 2026 Spotlight: Meta-Ctrl predicts robot plan failures before the LLM task planner writes them. Exact inference, no sampling. Llama 3 8B beats GPT-5.5 & Claude Opus 4.8 on VirtualHome. 🎤 Talk 8:30 am · 📍 Poster #104, 10:30–12 🧵

0:21
Edward Sun avatar
⟳ Guy Van den Broeck retweeted
Edward Sun@sebwarb1·Nov 11

Come check out our spotlight and poster talk today! 🎊

📄 CoRL 2026 Spotlight: Meta-Ctrl predicts robot plan failures before the LLM task planner writes them. Exact inference, no sampling. Llama 3 8B beats GPT-5.5 & Claude Opus 4.8 on VirtualHome. 🎤 Talk 8:30 am · 📍 Poster #104, 10:30–12 🧵

0:21
Gwen Yidou-Weng avatar
⟳ Guy Van den Broeck retweeted
Gwen Yidou-Weng@yidouweng·Nov 11

📄 CoRL 2026 Spotlight: Meta-Ctrl predicts robot plan failures before the LLM task planner writes them. Exact inference, no sampling. Llama 3 8B beats GPT-5.5 & Claude Opus 4.8 on VirtualHome. 🎤 Talk 8:30 am · 📍 Poster #104, 10:30–12 🧵

0:21
UCLA Computer Science avatar
⟳ Guy Van den Broeck retweeted
UCLA Computer Science@UCLAComSci·Nov 10

UCLA Computer Science is Hiring! The UCLA Computer Science Department at the Samueli School of Engineering invites applications for multiple tenure-track Assistant Professor positions in all areas of computer science. For more information and to apply: recruit.apo.ucla.edu/JPF1141…

Post attachment
Murat Kocaoglu avatar
⟳ Guy Van den Broeck retweeted
Murat Kocaoglu@murat_kocaoglu_·Oct 9

This is a very good take. Clarity is no longer coupled with performance in math and theory research. Both are important, and operating at this pareto front could be our goal. We just need to adjust academic incentives to reward the whole curve a bit better.

We're clearly moving to a pareto front of "strenght of result" vs. "understandability" and as long as you are on that pareto front, you are all good.

Guy Van den Broeck avatar
Guy Van den Broeck@guyvdb·Oct 8

We're clearly moving to a pareto front of "strenght of result" vs. "understandability" and as long as you are on that pareto front, you are all good.

Apparently, my coauthor pointed out that the OpenAI theorem drop included something we were working on (with AI help, of course). The OpenAI result has better bounds than we achieved, but it's also unreadable. Any advice on what I'm supposed to do now?

Talia Ringer 🕊🪬 avatar
⟳ Guy Van den Broeck retweeted
Talia Ringer 🕊🪬@TaliaRinger·Oct 6

Given the sudden and amazing ubiquity of formal proofs in systems that implement the calculus of inductive constructions these days, when will Thierry Coquand, Christine Paulin-Mohring, and Gérard Huet get the Turing Award?

Mike Knoop avatar
⟳ Guy Van den Broeck retweeted
Mike Knoop@mikeknoop·Oct 5

Cool research from @industriaalist. It's related to one of my favorite papers from @guyvdb arxiv.org/abs/2205.11502 which shows that transformers can represent many correct programs that backprop will never find. Below: a novel non-backprop based search mechanism.

Backprop has been the only credit assignment algorithm capable of training large neural nets. Introducing Dust, the first zeroth-order method to pretrain transformers to approach and sometimes even *exceed* backprop with large amounts of computation (bitter lesson!) - Dust is on the order of 1,000 to 10,000x more compute efficient than EGGROLL, the state-of-the-art ES method, for training transformers. - We chose pretraining transformers because it's arguably the hardest task possible for zeroth-order optimization. - Search in high dimensions is really misunderstood: zeroth-order usually gets better with overparameterization, not worse, i.e. at a fixed population size, larger models (up to 120x) reach a lower loss than smaller ones. - Backprop-like gradients emerge from a large population of activation perturbations, and the alignment holds up at every scale we tested, up to 1B tokens, which is critical for scaling. The core idea is a *virtual population*, which helps scale to large populations in transformers by perturbing activations in parallel. w/ @bishmdl76, @cs_serdar, @akshayvegesna

0:15
Aaron Roth avatar
⟳ Guy Van den Broeck retweeted
Aaron Roth@Aaroth·Sep 24

AI tools are very useful for mathematical research. But at present, producing good work requires a labor intensive step that I have been calling "deslopping" - making the thing readable. Please do not skip this step; even if the theorem is great an unreadable paper is worthless.

François Chollet avatar
⟳ Guy Van den Broeck retweeted
François Chollet@fchollet·Sep 8

Terence Tao just posted this really insightful thread which seems to have been written in reaction to the Navier Stokes announcement. Please give it a read. mathstodon.xyz/@tao/11723732…

F.Mackenzie 约克.小汽车. 嘟嘟 avatar
⟳ Guy Van den Broeck retweeted
F.Mackenzie 约克.小汽车. 嘟嘟@FMackenzie7·Sep 1

A Probabilistic Interpretation of KV Cache Eviction 🙇‍♂️🙇‍♂️🙇‍♂️🙇‍♂️🙇‍♂️

“A Probabilistic Interpretation of KV Cache Eviction” Most KV eviction methods use top-k heuristics to decide what to keep, but this can heavily distort attention after tokens are removed. This paper reframes KV eviction as expectation estimation, then uses probabilistic sampling and importance weighting to correct for missing KV entries during decoding. This gives a principled bias-variance tradeoff that stays competitive with existing methods while being more robust. alphaxiv.org/abs/2608.28293

Post attachment
Daniel Israel avatar
⟳ Guy Van den Broeck retweeted
Daniel Israel@danielmisrael·Sep 2

What happens if you interpret attention scores as probabilities? We can then use sampling techniques in KV cache eviction. Check out our new paper!

“A Probabilistic Interpretation of KV Cache Eviction” Most KV eviction methods use top-k heuristics to decide what to keep, but this can heavily distort attention after tokens are removed. This paper reframes KV eviction as expectation estimation, then uses probabilistic sampling and importance weighting to correct for missing KV entries during decoding. This gives a principled bias-variance tradeoff that stays competitive with existing methods while being more robust. alphaxiv.org/abs/2608.28293

Post attachment
alphaXiv avatar
⟳ Guy Van den Broeck retweeted
alphaXiv@askalphaxiv·Sep 1

“A Probabilistic Interpretation of KV Cache Eviction” Most KV eviction methods use top-k heuristics to decide what to keep, but this can heavily distort attention after tokens are removed. This paper reframes KV eviction as expectation estimation, then uses probabilistic sampling and importance weighting to correct for missing KV entries during decoding. This gives a principled bias-variance tradeoff that stays competitive with existing methods while being more robust. alphaxiv.org/abs/2608.28293

Post attachment