Fangzhi Xu (@leo_xu98) 's Twitter Profile
Fangzhi Xu

@leo_xu98

2nd Year PhD student @ XJTU

ID: 1739571718270889984

linkhttp://xufangzhi.github.io calendar_today26-12-2023 09:00:06

54 Tweet

87 Takipçi

152 Takip Edilen

Fangzhi Xu (@leo_xu98) 's Twitter Profile Photo

🥳Really excited to receive the interests from the community ! 🫡Many thanks to Maxi for building the demo for OS-Atlas ! Now everyone can have a try: huggingface.co/spaces/maxiw/O…

Fangzhi Xu (@leo_xu98) 's Twitter Profile Photo

🚀Important Updates: 13M GUI Grounding Pretraining Data is made public ! The largest grounding collection ever ! 🫡Check it: huggingface.co/datasets/OS-Co…

Fangzhi Xu (@leo_xu98) 's Twitter Profile Photo

🚀🥳Thrilled to release our recent efforts on bootstrapping VLMs to chain-of-thought reasoning ! arxiv.org/abs/2411.00855

Yu Gu @ICLR 2025 (@yugu_nlp) 's Twitter Profile Photo

❓Wondering how to scale inference-time compute with advanced planning for language agents? 🙋‍♂️Short answer: Using your LLM as a world model 💡More detailed answer: Using GPT-4o to predict the outcome of actions on a website can deliver strong performance with improved safety and

❓Wondering how to scale inference-time compute with advanced planning for language agents?

🙋‍♂️Short answer: Using your LLM as a world model
💡More detailed answer: Using GPT-4o to predict the outcome of actions on a website can deliver strong performance with improved safety and
Chengzu Li (@li_chengzu) 's Twitter Profile Photo

Forget just thinking in words. 🚀 New Era of Multimodal Reasoning🚨 🔍 Imagine While Reasoning in Space with MVoT Multimodal Visualization-of-Thought (MVoT) revolutionizes reasoning by generating visual "thoughts" that transform how AI thinks, reasons, and explains itself.

Forget just thinking in words.

🚀 New Era of Multimodal Reasoning🚨
🔍 Imagine While Reasoning in Space with MVoT

Multimodal Visualization-of-Thought (MVoT) revolutionizes reasoning by generating visual "thoughts" that transform how AI thinks, reasons, and explains itself.
Xu Huang (@xuhuang87) 's Twitter Profile Photo

🤩Excited to announce our new work BenchMAX!🥳 BenchMAX: A Comprehensive Multilingual Evaluation Suite for Large Language Models Paper: huggingface.co/papers/2502.07… Repo: github.com/CONE-MT/BenchM… Datasets: huggingface.co/collections/LL…

🤩Excited to announce our new work BenchMAX!🥳

BenchMAX: A Comprehensive Multilingual Evaluation Suite for Large Language Models

Paper: huggingface.co/papers/2502.07…
Repo: github.com/CONE-MT/BenchM…
Datasets: huggingface.co/collections/LL…
Fangzhi Xu (@leo_xu98) 's Twitter Profile Photo

🤩Excited to receive positive feedbacks from the community ! Please stay tuned with our another paper, named Genius, on applying foresight sampling on RL post-training !

Sansa Gong (@sansa19739319) 's Twitter Profile Photo

🤖Can diffusion models write code competitively? Excited to share our latest 7B coding diffusion LLM!!💻 With DiffuCoder, we explore how they decode, why temperature🔥 matters, and how to improve them via coupled-GRPO that speaks diffusion!!📈 Code: github.com/apple/ml-diffu… 🧵

🤖Can diffusion models write code competitively?
Excited to share our latest 7B coding diffusion LLM!!💻

With DiffuCoder, we explore how they decode, why temperature🔥 matters, and how to improve them via coupled-GRPO that speaks diffusion!!📈

Code: github.com/apple/ml-diffu…
🧵