Shanghai, China

Xuansheng Wu

Elite
@JacksonWuxs

I am a researcher at Shanghai AI Lab. I study how LLMs work and how we can make them better by understanding them better.

jacksonwuxs.github.io/

DaPy. Easy-to-use data analysis / manipulation framework for humans

590

Forward-Forward-Network. Implementation of Forward Forward Network proposed by Hinton in NIPS 2022.

170

UsableXAI_LLM. Using Explanations as a Tool for Advanced LLMs

70

PromptRec. Prompting Small Language Models for Personalized Cold-Start Recommendation

32

Interpret_Instruction_Tuning_LLMs. Understanding Why and How Instruction Tuning Changes Pre-trained Models

25

taac2021-Video-Classification-Rank5. 2021腾讯广告算法大赛-赛道二-第五名方案

20

19nCoV-SEIR-Estimation. An adjustive SEIR model to estimate parameters of 2019-nCoV

19

BeeDrive. BeeDrive: Open Source Privacy File Transfering System for Teams and Individual Developers

12

SelfReg-Achieving_Controllable_LLM_Classifier_with_SAE. Python

11

Syntactic-Attention-based-Random-Walk. a quick and high performance sentence embeddnig approach

7

MeNSP. Matching Examplers as Next Sentence Prediction

6

DIRECT. Dual Interpretable Recommendation with Multi-aspect Word Attribution

4

easyThreading. A user-friendly threading management library for humans

2

SteeringLLMs_with_MIExplanations. Interpreting and Steering LLMs with Mutual Information-based Explanations on Sparse Autoencoders

2

Flash-Background-by-wxPython. My sollution on flashing background in wxPython

1
15
Apply