허밍 lab

  • 홈
  • 태그
  • 방명록

mldl 2

[강의] CS25 Transformers United V4 3강: Aligning Open Language Models(Nathan Lambert)

강의 출처: https://web.stanford.edu/class/cs25/ CS25: Tranformers United!Disussing the latest breakthroughs with Transformers in diverse domainsweb.stanford.edu Aligning Open Language ModelsA heavily abbreviated history of language models (LMs)cf) stochastic parrots: 언어 모델이 단순히 데이터를 흉내 내는 역할만 한다는 비판적 시각https://dl.acm.org/doi/10.1145/3442188.3445922Can ChatGPT exist without RLHF?cf) RLHF(Reinforcemen..

learning/Lectures 2025.01.20

[강의] CS25 Transformers United V4 2강: Jason Wei & Hyung Won Chung of OpenAI

강의 출처: https://web.stanford.edu/class/cs25/Intuitions on Language Models (Jason)Q. Why do LMs work so well?→ manually inspect dataex) 폐암의 종류를 분류하는(classify) 프로젝트를 했었음but 이 일을 하려면 medical degree 필요하다고 함논문 많이 읽고, 병리학적 case 많이 접함How LM are trainednext word prediction task: 앞에 나오는 단어가 있고, 그 다음에 나올 단어를 예측함각 단어에 대한 확률 도출LM은 모든 단어에 확률 부여손실(loss): 실제 다음 단어와 예측 단어 간 차이→ 첫번째 직관*: next-word prediction = ma..

learning/Lectures 2025.01.20
이전
1
다음
더보기
프로필사진

허밍 lab

  • 분류 전체보기 (15)
    • thinking (0)
    • learning (15)
      • MLDL (6)
      • HCIAI (3)
      • Lectures (6)
      • Textbooks (0)
    • living (0)

Tag

CHI, GPT, NLP, affective computing, conversational AI, cs25, ML, Zero shot, Prompt Engineering, llm, 논문리뷰, HCI, Stanford, dl, mldl, Ai, 인공지능, Transformer, open ai, vlm,

최근글과 인기글

  • 최근글
  • 인기글

최근댓글

공지사항

페이스북 트위터 플러그인

  • Facebook
  • Twitter

Archives

Calendar

«   2026/02   »
일 월 화 수 목 금 토
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28

방문자수Total

  • Today :
  • Yesterday :

Copyright © AXZ Corp. All rights reserved.

티스토리툴바