🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Dinghuai Zhang 张鼎怀 Tim Xiao Weiyang Liu Yoshua Bengio Code released: github.com/lzzcd001/nabla…. Feel free to check it out!

The implementation of Nabla-GFlowNet (#ICLR2025) is released. Welcome to try it out. :)

📢Glad to introduce FormalMATH, a large-scale Lean4 benchmark comprising 5,560 formally verified problems. 📖The benchmark spans from high-school Olympiad challenges to undergraduate-level theorems across diverse domains. The best LLM prover only achieved 16.46% accuracy. 1/4

📣 Excited to share our #CVPR2025 Spotlight paper and my internship project Wayve: SimLingo. A Vision-Language-Action (VLA) model that achieves state-of-the-art driving performance with language capabilities. Code: github.com/RenzKa/simlingo Paper: arxiv.org/abs/2503.09594

Checkout our recent work on efficient pretraining for LLM!

I was surprised when I first saw that the black magic of prompt engineering can marry classical ML methods in such a natural way - simply asking an LLM to do rejection sampling makes it a more rational agent. Cannot wait to see how we may similarly design better "LLM algorithms".

Great paper by my students Tim Xiao and Johannes Zenn and collaborators that applies ideas from Monte Carlo sampling to (black-box) LLM execution to turn LLMs into better calibrated stochastic samplers.

Verbalized machine learning treats LLMs with prompts as function approximators. Building on this, Tim Xiao came up with the idea of studying whether LLMs can act as samplers. It turns out they’re often biased, even when they appear to understand the target distribution.

Muon is gaining attention for its use of orthogonalization, making it a natural point of comparison with POET. We computed singular value entropy over training steps and find that POET always maintains high entropy. A recent study (arxiv.org/abs/2502.16982) suggests that this is a

Our #ICCV2025 HANDS workshop will be on Oct. 20, PM! We focus on hand-related areas, e.g., hand pose est., hand-object interaction, robotics hand manipulation. hands-workshop.org NUS ETH CS Department Uni of Birmingham Reality Labs at Meta AI at Meta 東京大学 | UTokyo Meshcapade

We have added some new experiments and analyses to the new version of our paper. Check it out here: arxiv.org/abs/2506.08001. We discovered that despite being generalized to spectrum-preserving training, POET can still preserve minimum hyperspherical energy. This property only