AI Briefing#llm
The latest AI and developer news about #llm, with the original source and a short summary.
The latest AI and developer news about #llm, with the original source and a short summary.
AutoAdapt automates the selection and tuning of RAG and fine-tuning within given constraints.
A single checkpoint switches among 8 presets, delivering up to 10.7x faster performance.
Qwen 3.6 27B, a new model in Alibaba's Qwen series, has been released.
Introduces the "Skillify" methodology, which permanently fixes AI agent mistakes by converting them into markdown skills and deterministic scripts.
Anthropic appears to be moving to remove Claude Code access from the Pro plan and shift to API-based billing.
Opus 4.7 keeps the same price, but token counts can increase up to 1.35x.
Google has released LiteRT-LM, its on-device LLM inference engine.
Google unveiled Deep Research and Deep Research Max, built on Gemini 3.1 Pro to maximize performance and efficiency.
Apple unveiled research on RNNs, SSMs, multimodal, 3D, and protein folding at ICLR 2026.
Checklists and critique decomposition pushed the LLM Judge's agreement rate with humans up to 84%.
The problem of generating explanations with no ground-truth data is solved through policy consensus and iterative evaluation.
SpaceX announced a $60 billion acquisition agreement with Cursor.
IBM has released Granite-4.1-8B, enhancing tool calling and instruction-following performance.
A 4-bit reasoning model based on DeepSeek-R1-32B has been released, reduced to 20GB.
Unsloth has released a GGUF distribution of Kimi K2.6 along with related documentation.
Trained a 235M-parameter LLM from scratch using a single RTX 5080.
Lex Custis, an open-source compliance engine for EU AI Act compliance, has been released.
BAR trains domain-specific experts separately, merges them into one, and lets you swap out just a few.
GoModel has been released as an OpenAI-compatible gateway that bundles multiple LLM providers together.
By optimizing the q1_0/q8_0 dot path, up to 67x throughput improvement was reported on x86.
Amazon and Anthropic have signed an AI infrastructure agreement worth up to $125 billion over 10 years.
A major cloud deal has been revealed in which Amazon will invest up to $25 billion in Anthropic.
MoE models achieve higher inference acceleration when Speculative Decoding techniques are applied.