Yixin Song's picture

Yixin Song

yixinsong

·

AI & ML interests

None yet

Recent Activity

upvoted a paper 4 days ago

Demons in the Detail: On Implementing Load Balancing Loss for Training Specialized Mixture-of-Expert Models

updated a model 10 days ago

PowerInfer/SmallThinker-3B-Preview

new activity 10 days ago

PowerInfer/SmallThinker-3B-Preview:Eval script

View all activity

Organizations

yixinsong's activity

upvoted a paper 4 days ago

Demons in the Detail: On Implementing Load Balancing Loss for Training Specialized Mixture-of-Expert Models

Paper • 2501.11873 • Published 5 days ago • 59

upvoted a paper about 2 months ago

Densing Law of LLMs

Paper • 2412.04315 • Published Dec 5, 2024 • 17

upvoted a paper 4 months ago

Configurable Foundation Models: Building LLMs from a Modular Perspective

Paper • 2409.02877 • Published Sep 4, 2024 • 28

upvoted 3 papers 5 months ago

Dolphin: Long Context as a New Modality for Energy-Efficient On-Device Language Models

Paper • 2408.15518 • Published Aug 28, 2024 • 43

Scaling Synthetic Data Creation with 1,000,000,000 Personas

Paper • 2406.20094 • Published Jun 28, 2024 • 98

Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing

Paper • 2406.08464 • Published Jun 12, 2024 • 67

upvoted 2 papers 6 months ago

Q-Sparse: All Large Language Models can be Fully Sparsely-Activated

Paper • 2407.10969 • Published Jul 15, 2024 • 22

Qwen2 Technical Report

Paper • 2407.10671 • Published Jul 15, 2024 • 161

upvoted 3 papers 8 months ago

ReLU^2 Wins: Discovering Efficient Activation Functions for Sparse LLMs

Paper • 2402.03804 • Published Feb 6, 2024 • 2

PowerInfer-2: Fast Large Language Model Inference on a Smartphone

Paper • 2406.06282 • Published Jun 10, 2024 • 37

Turbo Sparse: Achieving LLM SOTA Performance with Minimal Activated Parameters

Paper • 2406.05955 • Published Jun 10, 2024 • 24

upvoted a paper 10 months ago

Physics of Language Models: Part 3.2, Knowledge Manipulation

Paper • 2309.14402 • Published Sep 25, 2023 • 7

upvoted a paper 11 months ago

The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits

Paper • 2402.17764 • Published Feb 27, 2024 • 607

upvoted a collection 12 months ago

BrockportGPT v2

The improved version of BrockportGPT v1. This generation has enhanced datasets that are more usable. See https://github.com/msaad02/honors-thesis • 7 items • Updated Mar 18, 2024 • 1

upvoted a paper about 1 year ago

PIA: Your Personalized Image Animator via Plug-and-Play Modules in Text-to-Image Models

Paper • 2312.13964 • Published Dec 21, 2023 • 19