AI Briefing#moe
The latest AI and developer news about #moe, with the original source and a short summary.
The latest AI and developer news about #moe, with the original source and a short summary.
Qwen3.6-27B, an open-source sparse MoE model optimized for agentic coding, has been released.
An abliteration version targeting the MoE characteristics of Qwen3.6-35B-A3B has been uploaded to HF.
The uncensored Qwen3.6-35B-A3B Aggressive version and K_P GGUF quants have been released.
Qwen3.6-35B-A3B has been released, achieving large-model-class coding and multimodal performance with just 3 billion active parameters.
Alibaba's Qwen team has released Qwen3.6-35B-A3B, an open-source MoE-based model with significantly improved agentic coding performance.
Qwen released a MoE model with 35B total parameters and 3B active parameters under Apache 2.0.
DFlash doesn't speed up as much as expected with MoE, quantization, concurrent use, and long context.
DeepSeek has updated DeepGEMM with Blackwell optimizations and Mega MoE support.
Built a paper-based web app with Qwen3.5 35B and confirmed it maintains context well up to 70-80K tokens.
An open-source book and repository implementing everything from GPT-2 to Llama 3.2-3B to DeepSeek in code.
LazyMoE aims for ultra-low-memory execution of a 120B LLM using lazy expert loading and TurboQuant.
Warp Decode shifts the parallelization axis of MoE inference from experts to outputs on Blackwell GPUs, boosting performance by 1.84x.
This explains how to run Claude Code in a local environment using LM Studio's new CLI and Gemma 4.
Google Gemma 4 26B MoE and 31B Dense are now available on Vercel AI Gateway.
AI21 has unveiled the **Jamba 1.5** family, a high-performance open model lineup based on an SSM-Transformer architecture.
By attaching CPT and BM25 RAG to Qwen3-30B-A3B, the team won the Excellence Award while running in a 24GB environment.
China's open-source AI ecosystem is evolving through MoE architecture, multimodal expansion, and small model optimization.
SAPO stabilizes LLM reinforcement learning by using smooth gating instead of hard clipping.
Mistral has announced the Mistral 3 series, its next-generation models released under the Apache 2.0 license.
Rebellions and SqueezeBits held a hands-on workshop where participants ran vLLM directly on NPUs.
Google Cloud C4 VMs based on Intel Xeon 6 improved TCO by 1.7x for GPT OSS inference.
Two RNGD cards served GPT-OSS 120B at 5.8ms TPOT with low power consumption.
OpenAI has released the GPT OSS series, an open-weight model lineup under the Apache 2.0 license.