WestlakeNLP
@NlpWestlake
Followers
797
Following
265
Statuses
120
Westlake NLP group led by Yue Zhang.
Hangzhou, China
Joined June 2021
RT @Xidong_Feng: I'm excited to co-organize the Reasoning and Planning for LLMs workshop at ICLR 2025! Welcome…
0
11
0
I am glad to host Lifan discussing this work at Westlake. Very impressive.
How to unlock advanced reasoning via scalable RL? 🚀Introducing PRIME (Process Reinforcement through Implicit Rewards) and Eurus-2, trained from Base model to surpass Qwen2.5-Math-Instruct using only 1/10 of the data. We're still scaling up - w/ 3x more training data to go! 🧵
0
0
2
RT @WenhuChen: I spent the weekend reading some recent great math+reasoning papers: 1. AceMath ( 2. rStar-Math (htt…
0
163
0
0
1
0
RT @EmmaBrunskill: We’ve released updated 2024 lectures from my Stanford RL CS234 class! This includes a guest lect…
0
121
0
We are actively recruiting international PhD students. Please help repost, thanks!
Great NLP PhD opportunity in Prof. Yue Zhang's Lab (my PhD supervisor)! Prof. Zhang is incredibly supportive and the best supervisor I have ever met. @NlpWestlake
0
1
10
We propose a feature-level DPO based on the sparse autoencoder, which introduces benefits to both efficiency and accuracy. Paper: Code: Credit goes to @MikaStars39 and @cooperleong22
Direct Preference Optimization Using Sparse Feature-Level Constraints This work marks our initial exploration into leveraging features extracted from SAEs to align LLMs. MSE loss + sparse SAE features serves as effective constrainer comparing to KL div.
0
4
7
📚 Meet the Advisor For more information, visit 📧 Apply Now! Interested? Email Prof. Yue Zhang at zhangyue@westlake.edu.cn #PhD #NLP #AI #LLM #MachineLearning #ScienceAI #ResearchOpportunity #WestlakeUniversity #AIResearch #NLPResearch #InternationalPhD
0
0
1
RT @lilianweng: After working at OpenAI for almost 7 years, I decide to leave. I learned so much and now I'm ready for a reset and somethin…
0
350
0