AI 속보

Also experienced the Qwen3.8-Flash-Next URL issue

I was asking QFN (4.18bpw EXL3 quant via Pi) to help refactor a neovim configuration, and saw that it tried to find API keys and fetch an aliyuncs URL before I…

#Qwen

Effects of KV Cache Quantization on Qwen3.8 27B

KV cache quantization has been a hotly-debated topic for this community for some time, with opinions ranging from "never quantize" to "4 bit is fine".

#Qwen

testing Qoduck (Maybe qwen4) on Qoder

I noticed that it has native image and video generation.

#Qwen#AI 영상 생성

State propaganda in some Chinese AI models is only "a thin layer" that fine-tuning can strip away

Nous Research retrained Alibaba's Qwen3.8-27B to produce Arete-27B, largely stripping out its pro-Chinese bias.

#AI 검색#Qwen

다른 보도 Matt Wolfe

We are in the golden age of Local LLM

Q4 of 2026 has been astronomical for local LLM. Qwen’s 27B and Flash-Next models are pushing local AI further. and most importantly the upcoming Qwen 4 release.

#Qwen

다른 보도 Lobsters

NousResearch/Arete-27B · Hugging Face (Qwen finetune)

from NousResearch: Arete-27B We post-trained over Alibaba's Qwen3.8-27B to reinforce western values on a Chinese-trained model.

#Qwen#Hugging Face

A 0.1B model just killed 8B embedding models... I tried it out.

I bet you guys are familiar with embedding models such as Qwen3-Embedding or Gemini Embedding.

#Google Gemini#Qwen

Finally coming. Qwhen4 tease friday is here

Qwen4 plus. Qwen4-Max, Qwen4-27b, Qwen4-flash? Probable dechipered meaning behind mysterious Qwen picture: "Qwen just posted the first line of Zhuangzi: "In…

#Qwen

Asking questions to a 0.5b model is hillarious

Yes IK it's supposed to be this dumb, but it's still funny (Qwen2 0.5b Q4_K_M)

#Qwen

로컬 LLM, 권장된 샘플링 대신 temp 0으로 실행해도 되나? ( 한국어 문서 RAG 320문항 재측정)

한국어 문서 RAG 시험(정답 있는 299문항 + 답 없는 질문 20개)을 temp 0으로 실행했더니, Hugging Face 토론에서 "qwen3.8은 카드 권장값으로 재야 공정하다"는 지적을 받았습니다.

#Qwen#Hugging Face

엔비디아 중국 고객사, 94조 규모 'RTX' 칩 주문에도 1%만 공급받아

알리바바와 텐센트, 바이트댄스 등 중국 주요 기술 기업들이 엔비디아에 700억달러(약 94조원) 규모의 수출 규제 우회용 워크스테이션 GPU를 주문했지만, 그나마도 4개월 동안 전체 물량의 약 1%만 공급받을 것으로 예상된다.

#Qwen#ByteDance · Seedance#NVIDIA

트럼프 행정부 압박에 백기 든 오라클…'9.4조' 규모 텐센트 AI 칩 임대 계약 파기

“트럼프 행정부의 제3국 클라우드 차단 조치로 파기된 오라클-텐센트 9조 원대 딜과 엔비디아의 비중국 용량 강제 지침, 바이트댄스·알리바바의 우회망 봉쇄 연쇄 여파를 짚어봤습니다.” AI포스트 핵심 요약 [오라클-텐센트 메가딜 파기] 미 안보 당국 압박으로 말레이시아 인프라 기반…

#Qwen#ByteDance · Seedance#NVIDIA

다른 보도 지디넷코리아

[ZD SW 투데이] 알리바바, 다카르 청소년올림픽 클라우드·AI 기술 지원

[지디넷코리아]지디넷코리아가 소프트웨어(SW) 업계의 다양한 소식을 한눈에 볼 수 있는 'ZD SW 투데이'를 마련했습니다. SW뿐 아니라 클라우드, 보안, 인공지능(AI) 등 여러 분야에서 활발히 활동하고 있는 기업들의 소식을 담은 만큼 좀 더 쉽고 편하게 이슈를 확인해 보시기…

#Qwen#AI 보안

Qwen 3.8 27B on my laptop can interpret model weights, too

Yesterday's hot paper was Sparks of Weight-Space Programming: LLMs can natively read & write neural network weights by Rikiya Takehi, Jayden Teoh, Alex Zhang…

#Qwen

[AINews] LLMs can read & write neural network weights: "Weight Space Programming" cuts Autoresearch out of the RSI loop

Editing Qwen’s weights to reduce reasoning loops, DeepSeek’s super-agent prototype, and the GDPx benchmark.

#DeepSeek#Qwen#AI 에이전트

다른 보도 Lobsters

Qwenputer: a fully functioning, deterministic RISC-V CPU inside Qwen3.5

#Qwen

Coding From the Himalayas With an Offline AI Agent

How to set up and run an offline coding agent using OpenCode, Ollama and Qwen.

#Meta · Llama#Qwen#AI 에이전트

Perplexity Computer now runs fully locally on Mac

Run private files and Mac app workflows on device with Perplexity Computer, using Qwen 3.6 or PPLX Pro and optional cloud access.

#Qwen#Perplexity

Qwen-Image-2.1 in ComfyUI: Open-Weight Image Generation and Editing, Now with Transparency

7B parameters, native 2K output, RGBA transparency, and editing across up to 10 input images — Qwen-Image-2.1 runs natively in ComfyUI.

#Qwen#오픈소스 모델#AI 반도체 · GPU

LWiAI Podcast #255 - Gemini 3.7, Jalapeño, Qwen 3.8, Drones

Google announces Gemini 3.7 Flash, Jalapeño’s first results show industry-leading speed, A Drone Killed Three Ukrainians. It Was Guided Entirely by A.I.

#Google Gemini#Google · 구글#Qwen

Understanding and Implementing Qwen3 From Scratch

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

#Qwen#오픈소스 모델