Guy Van den Broeck

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

[Metin Alp @alpmetin1]

Make sure to check out our paper!

[Edward Sun @sebwarb1]

Come check out our spotlight and poster talk today! 🎊

[Gwen Yidou-Weng @yidouweng]

📄 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 🧵

[UCLA Computer Science @UCLAComSci]

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: https://t.co/VgkjsqU7H6

[Murat Kocaoglu @murat_kocaoglu_]

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.

[Guy Van den Broeck @guyvdb]

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.

[Lev Reyzin @lreyzin]

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 🕊🪬 @TaliaRinger]

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 @mikeknoop]

Cool research from @industriaalist. It's related to one of my favorite papers from @guyvdb https://t.co/e9cPNyPhXe which shows that transformers can represent many correct programs that backprop will never find. Below: a novel non-backprop based search mechanism.

[Guy Van den Broeck @guyvdb]

@CsabaSzepesvari that's correct

[Aaron Roth @Aaroth]

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.

[ICLR @iclr_conf]

Making Google's Paper Assistant Tool available to ICLR submitters: https://t.co/x94M7aeZSg

[François Chollet @fchollet]

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. https://t.co/SiuLLyp8jQ

[F.Mackenzie 约克.小汽车. 嘟嘟 @FMackenzie7]

A Probabilistic Interpretation of KV Cache Eviction

🙇‍♂️🙇‍♂️🙇‍♂️🙇‍♂️🙇‍♂️

[Daniel Israel @danielmisrael]

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

[alphaXiv @askalphaxiv]

“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.

https://t.co/ef2vO1ZIw0

[Guy Van den Broeck @guyvdb]

@jhoangeth I don't believe so.

[Guy Van den Broeck @guyvdb]

This might be a good time to mention my recent keynotes titled: "Agentic AI is Neurosymbolic AI"

[François Chollet @fchollet]

I would have assumed it was fairly obvious, but in case it's not: a million-line codebase (also known as a "harness"), running at inference time, orchestrating thousands of calls to a neural network for any given task, is the exact definition of a "neurosymbolic architecture"

[François Chollet @fchollet]

I would have assumed it was fairly obvious, but in case it's not: a million-line codebase (also known as a "harness"), running at inference time, orchestrating thousands of calls to a neural network for any given task, is the exact definition of a "neurosymbolic architecture"

[antonio vergari ⚔️ @tetraduzione]

🚨 2yr postdoc position on #reliable #constrained generation with #deep #generative #AI with applications to biology 🚨

to work with me and @WouterBoomsmaDK on #AI that can generate complex structures guaranteed to satisfy physical constraints

👇👇👇
https://t.co/oq9A4IppMV