機械が「意味」を持つとはどういうことか -Embedding-
AI transforms simple numbers (tokens) into "semantic coordinates (vectors)." Similar meanings gather in similar locations, and semantic relationships can be calculated using addition and subtraction. This map—embedding—is explained using real word vectors. The AI mechanism is explained with familiar examples and diagrams. From intuition to mechanism, formulas, and implementation, each video gently delves into one topic. 📖 Timestamps 0:00 Opening — The Mystery of "Sweet" 1:09 Intuition: Meaning Can Be Represented by "Location" 2:27 Concrete Example: Real Word Vectors and Meaning Calculation 4:17 Mechanism: Embedding Matrix and Distribution Hypothesis, the Limit of 1 Word = 1 Point 6:21 Summary and Next Episode Preview ▶ Next Episode: What is AI Looking at Now to Respond? – Attention – 🔔 Subscribe to the channel so you don't miss the next episode → https://www.youtube.com/@zunda-tsumug... ── Cast ── 🟢 Zundamon: The Explainer. Likes to understand things in terms of structure. 🌶️ Kasukabe Tsumugi: The Listener. This video addresses all of the viewers' common questions. 【For those who:】 ・Use ChatGPT or Claude but don't know the inner workings ・Want to understand the mechanism rather than just the superficial usage 【Primary sources referenced】 Mikolov, Chen, Corrado, Dean 2013 "Efficient Estimation of Word Representations in Vector Space" (word2vec) https://arxiv.org/abs/1301.3781 Mikolov, Yih, Zweig 2013 "Linguistic Regularities in Continuous Space Word Representations" (king-man+woman analogy) https://aclanthology.org/N13-1090/ Pennington, Socher, Manning 2014 "GloVe: Global Vectors for Word Representation" https://aclanthology.org/D14-1162/ Peters et al. 2018 "Deep contextualized word "Representations" (ELMo) https://aclanthology.org/N18-1202/ Devlin, Chang, Lee, Toutanova 2018 "BERT" https://arxiv.org/abs/1810.04805 【BGM/SE】 BGM: Amacha Music Studio https://amachamusic.chagasi.com/ Sound Effect Lab https://soundeffect-lab.info/ 【Voice】 VOICEVOX: Zundamon, Kasukabe Tsumugi https://voicevox.hiroshiba.jp/ #AIexplanation #LLM #ChatGPT #LargeScaleLanguageModel #Zundamon

Nobody Breaks Celebrities Like Rowan Atkinson

Die Zombie-Simulation, die niemand erklären kann

LAWYER: If Cops Ask "Where Are You Coming From?" - Say These Words

The Strange Math That Predicts (Almost) Anything

Is RAG Still Needed? Choosing the Best Approach for LLMs

AIってなに? ②AIは「思考」しているの?【人工知能】ブラックボックスなニューラルネットワークを紐解く【解説】【ディープラーニング 深層学習】

How Transformers Achieve Parallel Computing: The Transformer Architecture Explained

Why China Built The World’s Most Unprofitable Train Network

Prof. Dr. Christian Bauckhage (Fraunhofer IAIS): KI - Wir haben noch gar nichts gesehen!

Transformers, the tech behind LLMs | Deep Learning Chapter 5

5 German things that Japanese people find unbelievable 😳 | Culture shock in Germany

Why AI doesn't see "words" – Tokenizers –

Terence Tao: Nobody Understands Why AI Actually Works

エントロピーの再発明 | 圧縮と知能 Part 1

ChatGPT, Gemini, Claude & Co erklärt: Wie Maschinen Sprache verstehen | Terra X Lesch & Co

God Says:"DON’T IGNORE THIS IMPORTANT LETTER I SENT YOU"/God Message Now/God Message
![[Zundamon's AI-related paper commentary #55] Latent Reasoning with Normalizing Flows](https://i.ytimg.com/vi/ar37mL4VwoA/hqdefault.jpg?sqp=-oaymwEjCNACELwBSFryq4qpAxUIARUAAAAAGAElAADIQj0AgKJDeAE=&rs=AOn4CLBGkJzs0u7EgbJpGRlweFodPgi0VQ)
[Zundamon's AI-related paper commentary #55] Latent Reasoning with Normalizing Flows

But what is the Fourier Transform? A visual introduction.

Mathe-News 🚨 KI löst das Erdős-Einheitsabstand-Problem!

