Heng Dong

💥Excited to share our paper “AlphaOne: Reasoning Models Thinking Slow and Fast at Test Time” at #EMNLP2025 🚀 this Friday, Nov. 7, during Gather Session 4. Come say hi virtually!👋 📄Paper: arxiv.org/pdf/2505.24863 🪩Website & Code: alphaone-project.github.io #AI #LLMs #Reasoning

Huge thanks for sharing our work @_akhaliq! AlphaOne deep dive & code release coming soon 🚀

@jyzhang1208 truly impressive work!

When we successfully built a framework that enables MLLM-based agents to plan for low-level manipulation tasks (a key component of EmbodiedBench), I was super excited! Could this be a step toward MLLM-based agents becoming so versatile that we no longer need dedicated VLA models?

🤔Tired of agents stumbling in the dark? Making the right decision is tough without complete information. 🔎Diffusion models are here to infer global information from local observations! ✈️arxiv: arxiv.org/abs/2410.13953 @MilindTambe_AI, @drdhxi, @Yanchen_Jiang, David Parkes.

Here is our paper openreview.net/pdf?id=q9jQPA6…

Present our novel work “Leveraging Hyperbolic Embeddings for Coarse-Fine Robot Design” in #ICLR2024 in #Vienna. Welcome to visit our website sites.google.com/view/hyperboli… if you are interested. Also many thanks to my excellent co-author @jyzhang1208. Without her, nothing could be done.

Our work in #ICML2023: Symmetry-Aware Robot Design with Structured Subgroups. sites.google.com/view/robot-des…

Very excited to share our #NeurIPS2022 paper: Low-Rank Modular Reinforcement Learning via Muscle Synergy. Motivated by human muscle synergy, we designed an efficient framework that achieves low-rank control over high DoF robots. Project: sites.google.com/view/synergy-r…