Accelerating Diffusion LLMs via Adaptive Parallel Decoding "We therefore introduce adaptive parallel decoding (APD), a novel method that dynamically adjusts the number of tokens sampled in parallel." "Notably, Dream with ADP surpasses the speed of autoregressive Qwen 7B and

Announcing the 2025 IJCAI Computers and Thought Award winner ✨Aditya Grover Aditya Grover, Inception UCLA. Dr. Grover is honored for uniting deep generative models, representation learning & RL to advance scientific reasoning. Congratulations! ijcai.org/awards

🚨 First Call for Participation – NeSy 2025 📍 Sept 8–10 | Santa Cruz, CA Join the longest-running conference on neurosymbolic AI! Our keynote speakers: Guy Van den Broeck , Thomas Kipf , Deborah McGuinness , Gary Marcus More info 👇

Had the pleasure of learning about TRACE by Gwen Yidou-Weng, Benjie Wang, and Guy Van den Broeck at ICML! It view alignment/controlled decoding through a Bayesian lens and derives a simple, principled, and effective new method. I highly recommend reading this paper!

Wish LM could plan—not just guess the next word? TRACE lets LM see all endings before each move. – Global control at inference time – Tractable lookahead via an HMM LM-proxy – Linear classifier per constraint Outperform RL, DPO, FUDGE—at just +20% decoding over base LM. #ICML2025

#Nesy2025 schedule is out! Keynote speakers include Guy Van den Broeck Thomas Kipf Deborah McGuinness and Gary Marcus. I'm looking forward to seeing what everyone's cooking and meeting everyone! NeSy 2025 2025.nesyconf.org/schedule/

It is almost time to welcome you all in Santa Cruz! 🦕 We will start with an exciting and timely keynote by Guy Van den Broeck on "Symbolic Reasoning in the Age of Large Language Models" 👀

Eleonora Giunchiglia Now, Guy Van den Broeck is giving the opening keynote arguing why symbolic AI is still relevant in the age of LLMs... With the help of Shrek!

Eleonora Giunchiglia Guy Van den Broeck How can reverend Bayes help us to incorporate constraints? With NeSy of course 👀 With applications in non-toxic LLM generation and safe AI driving! Guy Van den Broeck

Eleonora Giunchiglia Guy Van den Broeck Behind all of these very nice methods are one central trick... Circuits! ➕✖️ These are tractable generative neural networks 😍

Watch out PyTorch 👀👀 You got competition here. Awesome work and talk by Guy Van den Broeck

🔦Adaptive Parallel Decoding (APD) has been accepted as a spotlight paper at NeurIPS Conference ! I thank my collaborators, reviewers, and program organizers for this honor. A thread for those interested 🧵 (1/n)

"An hour of planning can save you 10 hours of doing." ✨📝 Planned Diffusion 📝 ✨ makes a plan before parallel dLLM generation. Planned Diffusion runs 1.2-1.8× faster than autoregressive and an order of magnitude faster than diffusion, while staying within 0.9–5% AR quality.

Diffusion 🤝 Autoregressive Fast high-quality generation

Plan autoregressively, denoise in parallel!

I gave a keynote at NeSy 2025 on "Symbolic Reasoning in the Age of Large Language Models" Check out the recording if you are curious about neurosymbolic generative AI: youtube.com/watch?v=OtmiJR…

Recordings of the NeSy 2025 keynotes are now available! 🎥 Check out insightful talks from Guy Van den Broeck , Thomas Kipf and Deborah McGuinness on our new Youtube channel. Topics include using symbolic reasoning for LLM, and object-centric representations youtube.com/@NeSyconference