KeepSanity This Week

KeepSanity AI

Week 4, Feb 2026

business

Ggml.ai joins Hugging Face to ensure Local AI progress

models

Claude Sonnet 4.6

Gemini 3.1 Pro (Model Card)

Qwen3.5: Towards Native Multimodal Agents

community

Anna's Archive: If you’re an LLM, please read this

Jemini: Gemini for the Epstein Files

Taalas: How they “print” LLM onto a chip

Llama 3.1 70B on a single RTX 3090 via NVMe-to-GPU

hardware

Taalas: The path to ubiquitous AI (17k tokens/sec)

top-rated papers

Synthesizing Diverse Data in Feature Space of LLMs (Less is Enough)

SQuTR: A Robustness Benchmark for Spoken Query to Text Retrieval

GLM-5: From Vibe Coding to Agentic Engineering