오늘의 짤방: If you get this, we can be friends via @MartinPiper
- 빅데이터/인공지능
- OpenAI: ChatGPT back in Italy after meeting watchdog demands
- Running Python micro-benchmarks using the ChatGPT Code Interpreter alpha
- DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation
- Visual Instruction Tuning - LLaVA: Large Language and Vision Assistant
- A model that uses human prompts and sketches to generate realistic fashion images
- QuiLLMan: Voice Chat with LLMs
- Meet PassGAN, the supposedly “terrifying” AI password cracker that’s mostly hype
- New prompt injection attack on ChatGPT web version. Markdown images can steal your chat data.
- Adversarial Machine Learning and Cybersecurity
- Twitter's Recommendation Algorithm (updated)
- 자율 AI 에이전트, ChatGPT 다음의 메가트렌드?
- Flyte: An Open Source Orchestrator for ML/AI Workflows
- flyte - Build production-grade data and ML workflows, hassle-free
- 과학책 번역가가 생성 AI 시대를 바라보는 방식
- YOLO9000: Better, Faster, Stronger
- How I Re-implemented PyTorch for WebGPU
- Secret Messages Can Hide in AI-Generated Media
- MTIA v1: Meta’s first-generation AI inference accelerator
- PyLLMs is a minimal Python library to connect to LLMs (OpenAI, Anthropic, Google, AI21, Cohere, Aleph Alpha, HuggingfaceHub) with a built-in model performance benchmark.
- CS25: Transformers United V2 Winter 2023(첫 강의
- NodePad is a simple LLM-assisted brainstorming experiment designed to help you quickly capture ideas, expand on them, question them, and organize them visually! Feel free to explore the app or check the User Guide right away 😎
- 데이터 기반 기업이 되기 위해 넘어야 하는 6가지 허들
- 애플의 뉴럴 엔진과 생성형 AI 접근법
- Data Mining vs Machine Learning: Choosing the Right Approach
- 🥞🦜 LLM Bootcamp 🦜🥞 Today, let's talk about LLMOps.
- How to secure Generative AI applications
- Writing my own ChatGPT Code Interpreter
- Palm2 has been leaked to be 340B params and trained on 3.6T tokens (7.4e24 FLOPs).
- Cargo Cult AI - Is the ability to think scientifically the defining essence of intelligence?
- Introducing the ChatGPT app for iOS
- 전사 AI 도입 vs 일단 보류··· 글로벌 CIO가 제안하는 최적의 생성형 AI 도입법
- PaLM 2 Technical Report
- 🐛Debug Mode🐛 `import langchain; langchain.debug=True`
- Donkeycar: a python self driving library
- ChatGPT: Your Personal Python Coding Mentor
- Generative AI course by Google
- Inside GitHub: Working with the LLMs behind GitHub Copilot
- ⭐️ Decoding OpenAI Evals
- Outperforming Specialized Vector Databases with MyScale
- DreamFusion: Text-to-3D using 2D Diffusion
- Google’s newest A.I. model uses nearly five times more text data for training than its predecessor
- Numbers every LLM Developer should know
- AI-powered coding, free of charge with Colab
- VAR이 선수 훈련 도구로···AI로 완전히 바뀐 테니스 대회의 코칭
- Google I/O: Ubiquitous AI가 SEO에 큰 변화를 가져올 것 (demandsphere.com)
- GPflow is a package for building Gaussian process models in Python.
- Multi-GPU Training on a single GPU System in 3 Minutes
- Scale entanglement
- Large-scale Near-deduplication Behind BigCode
- Simplify Your Machine Learning Projects
- Search Personalization at Netflix
- Texas A&M commerce professor fails entire class of seniors blocking them from graduating- claiming they all use “Chat GTP”
- Rebuff is designed to protect AI applications from prompt injection (PI) attacks through a multi-layered defense.
- Towards Expert-Level Medical Question Answering with Large Language Models
- Auto-Evaluation of Anthropic 100k Context Window
- The Practical Guides for Large Language Models
- Pre-Training to Learn in Context
- Hugging Face Releases StarCoder, the Next-Generation LLM for Seamless Code Generation
- StructGPT: A General Framework for Large Language Model to Reason over Structured Data
- StructGPT: A general framework for Large Language Model to Reason on Structured Data (Code)
- [독서광] 인공지능 소프트웨어 품질 보증을 위한 테스트 기법(본인이 쓴 글)
- ‘한국말 잘 하는’ 외산AI...네카오 “골든타임 놓칠라” 비상 대응 체제
- 챗GPT, 서드파티 플러그인 기능 공개… 유료 사용자 대상 베타 서비스
- “규제 산업일수록 오픈소스 AI 모델 필요”… 개방형 AI 만드는 투게더, 2000만 달러 투자 유치
- 빙·엣지 이번엔 다를까?··· 마이크로소프트의 재도전에 쏠리는 관심
- You probably don't know how to do Prompt Engineering
- Github Copilot and ChatGPT alternatives
- We can adopt techniques like prompt ensembles to improve LLM reliability. 🧵 [1/10]
- CodeT5+: Open Code Large Language Models for Code Understanding and Generation
- Text Classification via Large Language Models
- Why Most Startups Are Doing Data-Driven Decision Making Wrong
- The Artificial Intelligence Act by EU
- Guidance enables you to control modern language models more effectively and efficiently than traditional prompting or chaining. (by Microsoft)
- Random(Random(Random(Random())))
- LMQL - A query language for programming (large) language models.
- Explore what's possible with some example applications using LLM (by OpenAI)
- GitHub Copilot Chat 전체 프롬프트 유출 (twitter.com/marvinvonhagen)
- Google Bard API - The python package that returns response of Google Bard through API.
- Brex's Prompt Engineering Guide
- privateGPT - Ask questions to your documents without an internet connection, using the power of LLMs. 100% private, no data leaves your execution environment at any point.
- The Prize Is Won; The Simplest Universal Turing Machine Is Proved
- How to run Llama 13B with a 6GB graphics card
- A Former Pilot On Why Autonomous Vehicles Are So Risky
- Hummingbird is a library for compiling trained traditional ML models into tensor computations.
- OpenLM - 여러 LLM 호출이 가능한 OpenAI 호환 파이썬 클라이언트 (github.com/r2d4)
- DeepFaceLab: DeepFaceLab is the leading software for creating deepfakes.
- Google Previews Studio Bot, a Coding Bot for Android Development
- Here are four prompting techniques for solving difficult, multi-step problems with LLMs… 🧵 [1/8]
- nanoGPT - The simplest, fastest repository for training/finetuning medium-sized GPTs.
- New AI advancements drive Meta’s ads system performance and efficiency
- LLM Prompting 관련 각 회사의 문서 목록 및 입문자용 추천 강좌 목록
- LLM Bootcamp - Spring 2023
- PUBLIC -- Claude 100K Test
- Animated Drawings by Meta
- Creating Scalable OpenAI GPT Applications in Java
- ImageBind - Meta의 멀티모달 AI모델 (imagebind.metademolab.com)
- “대중화 한 발짝” 마이크로소프트, 코파일럿 지원 앱 및 테스트 규모 확대
- Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Models∗
- Evaluating Embedding APIs for Information Retrieval
- Here are 8 things impossible on ChatGPT but that Bard can do (for free):
- Performance Tips Using Postgres and pgvector
- Running Large Language Models Natively on Mobile and Laptops
- Introducing 100K Context Windows
- Recommender Systems: A Primer
- 네이버, 2023년 1분기 매출액 2조 2,804억 원 기록… “하이퍼클로바X 네이버 서비스 전반에 적용할 것”
- AI가 만든 웹사이트들이 광고 예산을 거덜내고 있다 (adlingo.org)
- "실수 없는 AI는 없다"…딥페이크·실수·편견은 불가피하지만 통제 가능
- 챗GPT로 직원을 대체하는 것은 ‘루저’의 생각
- “비즈니스 가치 도출 중” 중고차 유통 기업 카맥스의 GPT-3.5 활용기
- 과장 광고된 AI를 판별하라···‘AI 워싱’을 피할 수 있는 9가지 전략
- “챗GPT가 AI 투자 증가 촉진했다”··· 가트너, 조직 70%가 생성형 AI 사용 중
- 오르카, 클라우드 앱 보안 플랫폼에 GPT-4 통합
- “세계에서 가장 정확한 번역기” 딥엘, 보안 강화한 기업용 서비스로 한국 본격 진출
- AI도 헌법같은 보편적 원칙 필요··· 앤트로픽, ‘클로드 헌법’ 전문 공개
- 건초더미에서 바늘 찾기를 ‘몇 분만에’··· 다우 케미컬의 화학물질 탐색 AI 혁신 사례
- “AI로 인한 고용 불안보다 활용 기대 더 높아”… 마이크로소프트, 업무동향지표 2023 발표
- mpt-7b-storywriter - A 7B parameter LLM fine-tuned to support contexts with more than 65K tokens
- #GoogleIO2023 Keynote #sketchnotes
- Excited to announce our multi-modal agent API that can automatically chain 100,000+ of HF models (stable diffusion, whisper, OpenAssistant,...) for text, audio, image, video, time-series,… based on your commands.
- Chatbot Arena Leaderboard Updates (Week 2)
- WHY-WHAT-HOW Questions Regarding AI Safety
- FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance
- MusicLM: Generating Music From Text
- PaLM 2 Technical Report
- Diffusion Explainer - Learn how Stable Diffusion transforms your text prompt into image!
- Diffusion-Explainer source code
- Diffusion Explainer: Visual Explanation for Text-to-image Stable Diffusion
- BackgroundRemover - 이미지/영상에서 배경지우기 오픈소스 (github.com/nadermx)
- Databerry - 나만의 LLM Agent를 만드는 노코드 플랫폼 (github.com/gmpetrov)
- Office Overachievers Won't Be Happy About ChatGPT, Study Says
- 독일산 AI 번역 ‘딥엘’ “한국어 번역 수요 늘어”...8월 유료서비스 출시
- Language models can explain neurons in language models
- Language models can explain neurons in language models (paper)
- 구글, 타사에 도움 되는 AI 논문 발표 금지
- MPT-7B, 상업적으로 이용 가능한 오픈소스 LLM (mosaicml.com)
- Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting
- 💬 RasaGPT is the first headless LLM chatbot platform built on top of Rasa and Langchain.
- Understanding the Transformer architecture for neural networks
- Open LLMs - These LLMs are all licensed for commercial use
- 마이크로소프트, 특이점 오면 세계 1위 될까?
- DeepSpeed empowers ChatGPT-like model training with a single click, offering 15x speedup over SOTA RLHF systems with unprecedented cost reduction at all scales; learn how.
- StarCoder, BigCode 프로젝트가 공개한 코드용 LLM (huggingface.co)
- The Meeting of the Minds That Launched AI - There’s more to this group photo from a 1956 AI workshop than you’d think
- All Machine Learning Algorithms You Should Know in 2022
- The Only Probability Cheatsheet You'll Ever Need
- Machine learning is famous for its open resources. We collected 3 free courses about ML:
- Learn AI with GitHub Copilot for FREE from Microsoft.
- 무신사가 카테고리숍 추천을 하는 방법
- Distilling Step-by-Step! Outperforming Larger Language Models with Less Training Data and Smaller Model Sizes
- Q&A code
- LlamaIndex Blog - The official blog of LlamaIndex. We’ll be featuring release updates, guides, community showcases, and more!
- How to Trick Your Brain into Remembering Almost Anything
- “나만의 상상이 작품이 된다” AI 이미지 생성기 완벽 입문 가이드 (itworld.co.kr)
- BackgroundRemover is a command line tool to remove background from image and video using AI, made by nadermx to power https://BackgroundRemoverAI.com.
- LangChain makes it easy to assemble LLM components (e.g., models and retrievers) into chains that support question-answering
- I am an evaluation tool for question-answering built on LangChain.
- AI Incident Database (incidentdatabase.ai)
- AI 인프라 비용 딜레마의 현재와 미래
- LLaMA.go - go version of LLaMA
- Helsinki University makes Parkinson's disease breakthrough
- Using OpenAI with JavaScript
- Transformer to LLaMA by 강필성
- A simple/effective way to improve information retrieval is to break it into two stages
- 하드웨어
- CAN Injection: keyless car theft
- 고속 충전이 휴대폰 배터리를 망가뜨리지 않는 이유
- The Great CPU Stagnation
- “칩 회사가 이걸 만들었다고?” 인텔의 기묘한 실패작 12가지
- The Hertella Coffee Machine Mounted on a Volkswagen Dashboard (1959): The Most European Car Accessory Ever Made
- ARM or x86? ISA Doesn’t Matter
- Electronics Lab Bench Setup Guide
- The Backbone - This page provides a single entry point to all the posts in the story of how the world got online.
- computers are bad - newletter
- How Steve Jobs saved Apple with the iMac 25 years ago
- 기계의 기능은 많을수록 좋은가? : ‘바보폰’의 귀환
- 블록체인과 메타버스
- 읽을거리
보너스: The 12 types of ML papers via @MIT_CSAIL
댓글 없음:
댓글 쓰기