최신 글

자신을 표현하는 새로운 방법: Gemini가 이제 음악을 만들 수 있습니다

A new way to express yourself: Gemini can now create music

The Gemini app now features our most advanced music generation model Lyria 3, empowering anyone to make 30-second tracks using text or images.

구글 딥마인드의 인도 파트너십: 과학과 교육에서의 AI 규모 확대

Google DeepMind Partnerships in India: scaling AI in science and education — Google DeepMind

Google DeepMind brings National Partnerships for AI initiative to India, scaling AI for science and education

Import AI 445: 초지능의 시점; AI가 최전선 수학 증명을 해결; 새로운 머신러닝 연구 벤치마크

Import AI 445: Timing superintelligence; AIs solve frontier math proofs; a new ML research benchmark

Will 2026 be looked back on as the pivotal year for making decisions about the singularity?

Gemini 3 Deep Think: 과학, 연구, 엔지니어링 발전

Gemini 3 Deep Think: Advancing science, research and engineering

Our most specialized reasoning mode is now updated to solve modern science, research and engineering challenges.

Gemini Deep Think: 과학 연구의 미래를 재정의하다 — Google DeepMind

Gemini Deep Think: Redefining the Future of Scientific Research — Google DeepMind

Research papers point to the growing impact of Deep Think across fields

Import AI 444: LLM 사회; 화웨이의 AI 커널; 칩벤치

Import AI 444: LLM societies; Huawei makes kernels with AI; ChipBench

How can you quantify creativity?

Release v0.1.0 - Initial Release · rohitg00/awesome-claude-code-toolkit · GitHub

<h1>awesome-claude-code-toolkit v0.1.0</h1> <p>The most comprehensive toolkit for Claude Code.</p> <h2>What's Included</h2> <table> <thead> <tr> <th>Category</th> <th>Count</th> </…

Import AI 443: 안개 속으로: 몰트북, 에이전트 생태계, 그리고 전환기의 인터넷

Import AI 443: Into the mist: Moltbook, agent ecologies, and the internet in transition

Plus, a story about agents corrupting other agents

프로젝트 지니: 미국 Ultra 사용자를 위한 AI 월드 모델 공개

Project Genie: AI world model now available for Ultra users in U.S.

Google AI Ultra subscribers in the U.S. can try out Project Genie, an experimental research prototype that lets you create and explore worlds.

Import AI 442: AI 경제의 승자와 패자, 수학 증명 자동화, 그리고 사이버 첩보의 산업화

Import AI 442: Winners and losers in the AI economy; math proof automation; and industrialization of cyber espionage

Is superintelligence a phase change or a gradual shift?

LLM 추론 개선을 위한 추론 시간 스케일링의 카테고리

Categories of Inference-Time Scaling for Improved LLM Reasoning

And an Overview of Recent Inference-Scaling Papers

Import AI 441: 내 에이전트는 작동 중이야. 너의는?

Import AI 441: My agents are working. Are yours?

Plus: Corrupting AI systems with a poison fountain

D4RT: 통합 고속 4D 장면 재구성 및 추적 – Google DeepMind

D4RT: Unified, Fast 4D Scene Reconstruction & Tracking — Google DeepMind

D4RT: Unified, efficient 4D reconstruction and tracking up to 300x faster than prior methods.

Veo 3.1 재료에서 비디오로: 더 나은 일관성, 창의성, 제어

Veo 3.1 Ingredients to Video: More consistency, creativity and control

Our latest Veo update generates lively, dynamic clips that feel natural and engaging — and supports vertical video generation.

임포트 AI 440: 레드퀸 AI; AI 규제 AI; O-링 자동화

Import AI 440: Red queen AI; AI regulating AI; o-ring automation

How many of your are LLMs?

2025년 LLM의 현황: 진전, 문제, 그리고 예측

The State Of LLMs 2025: Progress, Problems, and Predictions

A 2025 review of large language models, from DeepSeek R1 and RLVR to inference-time scaling, benchmarks, architectures, and predictions for 2026.

LLM 연구논문: 2025년 목록 (7월~12월)

LLM Research Papers: The 2025 List (July to December)

In June, I shared a bonus article with my curated and bookmarked research paper lists to the paid subscribers who make this Substack possible.

2025 연간 회고

2025 Year in Review

An eventful year of progress in health and career, while making time for travel and reflection.

DeepSeek V3에서 V3.2로: 아키텍처, 희소 주의, 강화학습 업데이트

From DeepSeek V3 to V3.2: Architecture, Sparse Attention, and RL Updates

Understanding How DeepSeek's Flagship Open-Weight Models Evolved

세 가지 간단한 단계로 제품 평가하기

Product Evals in Three Simple Steps

Label some data, align LLM-evaluators, and run the eval harness with each change.

표준 LLM을 넘어서 - Sebastian Raschka 박사

Beyond Standard LLMs - by Sebastian Raschka, PhD

Linear Attention Hybrids, Text Diffusion, Code World Models, and Small Recursive Transformers

새로운 Principal 기술 IC들을 위한 조언: 나에게 쓰는 노트

Advice for New Principal Tech ICs (i.e., Notes to Myself)

Based on what I've learned from role models and mentors in Amazon

LLM 평가의 4가지 주요 접근법 이해하기 (기초부터)

Understanding the 4 Main Approaches to LLM Evaluation (From Scratch)

Multiple-Choice Benchmarks, Verifiers, Leaderboards, and LLM Judges with Code Examples

의미 ID를 활용한 제어 가능 추천을 위한 LLM-RecSys 하이브리드 훈련

Training an LLM-RecSys Hybrid for Steerable Recs with Semantic IDs

An LLM that can converse in English & item IDs, and make recommendations w/o retrieval or tools.

Qwen3을 처음부터 이해하고 구현하기

Understanding and Implementing Qwen3 From Scratch

A Detailed Look at One of the Leading Open-Source LLMs

GPT-2에서 gpt-oss로: 아키텍처 발전 분석

From GPT-2 to gpt-oss: Analyzing the Architectural Advances

And How They Stack Up Against Qwen3

주요 LLM 아키텍처 비교

The Big LLM Architecture Comparison

From DeepSeek-V3 to Kimi K2: A Look At Modern LLM Architecture Design

LLM 연구 논문: 2025년 목록 (1월~6월)

LLM Research Papers: The 2025 List (January to June)

A topic-organized collection of 200+ LLM research papers from 2025

긴 맥락 질의응답 시스템 평가

Evaluating Long-Context Question & Answer Systems

Evaluation metrics, how to build eval datasets, eval methodology, and a review of several benchmarks.

LLM의 KV 캐시 이해와 처음부터 구현하기

Understanding and Coding the KV Cache in LLMs from Scratch

KV caches are one of the most critical techniques for efficient inference in LLMs in production.