DEV Community

[画像:Jimin Lee profile picture]

Jimin Lee

Head of Language AI Part @ Samsung | AI/ML | Galaxy AI · Bixby | LLM · RAG · Machine Translation · NLU · Voice Assistant

Joined on Sep 13, 2025

22 posts published

0 comments written

0 tags followed

Unraveling AI Hallucinations: Why LLMs Lie and How to Tame Them

jiminlee profile

Unraveling AI Hallucinations: Why LLMs Lie and How to Tame Them

#llm #deeplearning #ai #hallucination

11 min read

PyTorch Tensor Internals: Storage, Shape, and Stride Explained

jiminlee profile

PyTorch Tensor Internals: Storage, Shape, and Stride Explained

#deeplearning #ai #pytorch

8 min read

TPU: Why Google Doesn’t Wait in Line for NVIDIA GPUs (2/2)

jiminlee profile

TPU: Why Google Doesn’t Wait in Line for NVIDIA GPUs (2/2)

#llm #tpu #deeplearning #ai

7 min read

TPU: Why Google Doesn't Wait in Line for NVIDIA GPUs (1/2)

jiminlee profile

TPU: Why Google Doesn't Wait in Line for NVIDIA GPUs (1/2)

#tpu #llm #deeplearning

8 min read

Demystifying loss.backward(): How PyTorch Autograd Actually Works

jiminlee profile

Demystifying loss.backward(): How PyTorch Autograd Actually Works

#deeplearning #pytorch #python

10 min read

Why GPUs Ate the AI World

jiminlee profile

Why GPUs Ate the AI World

#llm #deeplearning #nlp #gpu

8 min read

Giving AI Eyes: Multi-Modal LLMs

jiminlee profile

Giving AI Eyes: Multi-Modal LLMs

#llm #deeplearning #nlp

9 min read

Speculative Decoding: Making LLMs Faster Without Sacrificing Quality

jiminlee profile

Speculative Decoding: Making LLMs Faster Without Sacrificing Quality

#llm #ai #deeplearning #nlp

14 min read

Tokens: The Invisible Building Blocks of Large Language Models

jiminlee profile

Tokens: The Invisible Building Blocks of Large Language Models

#llm #token #nlp

12 min read

From Parrot to Partner - How Reinforcement Learning Taught LLMs to Talk Like Humans

jiminlee profile

From Parrot to Partner - How Reinforcement Learning Taught LLMs to Talk Like Humans

#llm #nlp #rl #alignment

13 min read

Understanding the KV Cache (feat. Self-Attention)

jiminlee profile

Understanding the KV Cache (feat. Self-Attention)

#llm #nlp #kvcache

12 min read

Matryoshka Magic: How MatFormer Delivers Multi-Size Models from a Single Training

jiminlee profile

Matryoshka Magic: How MatFormer Delivers Multi-Size Models from a Single Training

#llm #nlp #matformer

8 min read

Why Are LLMs So Slow? And How We're Making Them Faster

jiminlee profile

Why Are LLMs So Slow? And How We're Making Them Faster

11 min read

jiminlee profile

On-Device LLM

#llm #machinelearning #nlp #ondeviceai

10 min read

Understanding Context Window Size in LLMs

jiminlee profile

Understanding Context Window Size in LLMs

#llm #machinelearning

4 min read

jiminlee profile

RAG Explained

#llm #machinelearning #rag

6 min read

What Is LoRA? The Guide to Low-Rank Adaptation

jiminlee profile

What Is LoRA? The Guide to Low-Rank Adaptation

#llm #machinelearning #lora #nlp

8 min read

Understanding Mixture of Experts (MoE)

jiminlee profile

Understanding Mixture of Experts (MoE)

#llm #machinelearning #nlp #moe

5 min read

From Word Predictor to Thinking Partner: The Rise of Thinking Models

jiminlee profile

From Word Predictor to Thinking Partner: The Rise of Thinking Models

#ai #llm #machinelearning

7 min read

(3/3) LLM: In-Context Learning, Hype, and the Road Ahead

jiminlee profile

(3/3) LLM: In-Context Learning, Hype, and the Road Ahead

#llm #machinelearning

5 min read

(2/3) LLM: Data, Transformers, and Relentless Compute

jiminlee profile

(2/3) LLM: Data, Transformers, and Relentless Compute

#data #ai #machinelearning #llm

11 min read

(1/3) LLM: How LLMs Became the Bedrock of Modern AI

jiminlee profile

(1/3) LLM: How LLMs Became the Bedrock of Modern AI

#llm #machinelearning

4 min read

loading...