#efficient-ai

AI TCH

YO YouShin kim

Pine AI — PreAct: 실행 속도를 8.5~13배 높이는 ‘스마트 반복 학습’ 컴퓨터 사용 에이전트 가이드

최근 AI 에이전트는 화면을 보고 클릭하며 타이핑하는 등 사람처럼 컴퓨터를 다루는 수준까지 발전했습니다. 하지만 치명적인 문제가 있습니다. 똑같은 작업을 내일 다시 시킨다면, 에이전트는 처음부터 다시 화면을 읽고 추론하며 비싼 언어 모델(LLM)…

Jun 21, 2026 · 6 min read · 0 claps #preact #ai-agent #computer-use

AI SPT TCH

YO YouShin kim

MIND LAB — δ-mem: Large Language Models를 위한 효율적인 온라인 연성 기억(Associative Memory) 메커니즘

1. Executive Summary & Background

May 27, 2026 · 6 min read · 0 claps #llm #memorymechanism #associative-memory

AI

SU Sumit Sah

Revolutionizing Real-Time Video Generation: Meet Q-ARVD, The Quantization Breakthrough for…

How researchers solved the unique challenges of quantizing autoregressive video diffusion models to make real-time AI video practical.

May 22, 2026 · 4 min read · 0 claps #ai-video-generation #model-quantization #diffusion-models

AI MDA TCH

AY Ayushi Jain

Everything was working perfectly with AI… until someone asked it to draw a dog.

When AI came — especially tools like Codex, Claude — where it can access all your code directories, create new files, execute code… you…

Apr 21, 2026 · 3 min read · 0 claps #ai-coding-time-reduction #efficient-ai

AI HUM

YO YouShin kim

Toward Efficient Agents: A Survey of Memory, Tool learning, and Planning

Background and Overview

Jan 28, 2026 · 6 min read · 0 claps #ai-agent #llm-efficiency #aiops

HUM AI

AR Argha Paul

Bigger is not always better : The Shift to AI Efficiency

AI efficiency is now taking center stage as the era of simply building bigger models reaches its natural limits. While Large Language…

Jan 18, 2026 · 8 min read · 0 claps #artificalintelligence #llm #efficient-ai

AI SCI

HA Haze News

Project χθos: A Proof of Concept for a New Paradigm in Efficient AI

1. The Premise: Challenging the “Scale or Die” Dogma

Jan 07, 2026 · 7 min read · 0 claps #efficient-ai #new-paradigm #proof-of-concept

AI ART ECO LIF

TA Tailored Techworks

The “Divorce Clause”: Why You Must Be Ready to Fire Your Model Provider

In a volatile market, “Model Loyalty” is a strategic error. Here is how to architect for independence.

Dec 29, 2025 · 2 min read · 0 claps #ai #ai-adoption #efficient-ai

AI MDA

RA RAKTIM SINGH · Stackademic

Reasoning Compression: Reflection Distillation & Long-Context Compaction

How to get more thinking per token — for faster, cheaper, greener, and more reliable AI

Oct 28, 2025 · 6 min read · 12 claps #enterprise-ai #aireasoning #efficient-ai

AI TCH

MO Mohit Sewak, Ph.D. · Sharing Science

Flow Matching (The Next Gen AI Algorithm): From 1000-Steps to a Single-Step Media Generation…

How a simulation-free training method is quietly outpacing today’s state-of-the-art diffusion and normalizing flows models.

Oct 22, 2025 · 13 min read · 120 claps #generative-ai #diffusion-models #ai-image-generation

HUM AI

YO YouShin kim

초소형 네트워크를 활용한 재귀적 추론 (Less is More: Recursive Reasoning with Tiny Networks) 상세 분석

### 배경 및 개요

Oct 17, 2025 · 9 min read · 0 claps #llm #deep-learning #samsung-ai

AI

DR Dr. Shouke Wei

Unleashing Large Language Models with oLLM: Power on a Budget

Efficient AI Inference for Everyone, Anywhere

Oct 09, 2025 · 3 min read · 20 claps paywalled #llm #ollm #efficient-ai

HUM AI

YO YouShin kim

여러 심판관으로부터 효율적인 다중 턴 대화 평가자 학습하기 (Learning an Efficient Multi-Turn Dialogue Evaluator from Multiple…

배경 및 개요

Aug 10, 2025 · 8 min read · 0 claps #llm-as-a-judge #reward-model #multiturndialogue

AI

RT R. Thompson (PhD) · AI Simplified in Plain English

The Grok 3 Effect: How Mixture of Experts Models Are Reshaping AI

“We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten.” —…

Jul 31, 2025 · 3 min read · 0 claps paywalled #grok-3 #sparse-models #mixture-of-experts

ART HUM MDA AI LIF

RT R. Thompson (PhD)

The Future of AI is Small: Slim Architectures and the Push for Efficient On-Device ML

Machine learning has traditionally been a field of expansion, with researchers pushing the boundaries by building ever-larger models. From…

Mar 18, 2025 · 5 min read · 0 claps paywalled #slim-ai #edge-ml #efficient-ai