<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>AI/ML Field Guide — news</title><description>Daily AI and machine learning news briefs in plain English, with sources.</description><link>https://www.aimlfieldguide.com/</link><language>en-gb</language><item><title>vLLM v0.33.0 ships with 927 commits as v0.34.0 release candidates land</title><link>https://www.aimlfieldguide.com/news/2026-11-12-vllm-v0-33-0-ships-with-927-commits-as-v0-34-0-release/</link><guid isPermaLink="true">https://www.aimlfieldguide.com/news/2026-11-12-vllm-v0-33-0-ships-with-927-commits-as-v0-34-0-release/</guid><description>vLLM tagged v0.33.0 on 2 November 2026 with shorter Docker image names, then published two v0.34.0 release candidates within ten days.</description><pubDate>Thu, 12 Nov 2026 00:00:00 GMT</pubDate><category>mlops-and-deployment</category><category>efficient-ml</category><category>open-source-ml</category><category>opensource</category></item><item><title>Anthropic launches IPO roadshow, offering 825 million Class A shares at $70 to $78 on Nasdaq</title><link>https://www.aimlfieldguide.com/news/2026-11-10-anthropic-ipo-roadshow/</link><guid isPermaLink="true">https://www.aimlfieldguide.com/news/2026-11-10-anthropic-ipo-roadshow/</guid><description>The proposed listing under the ticker ANT follows the S-1 filed on 26 October. Anthropic itself is selling 801 million shares, with no proceeds, pricing date or closing date yet stated.</description><pubDate>Tue, 10 Nov 2026 00:00:00 GMT</pubDate><category>careers-and-industry</category><category>anthropic</category></item><item><title>Hugging Face adds a decision pipeline for models that score options instead of writing text</title><link>https://www.aimlfieldguide.com/news/2026-11-10-hugging-face-adds-a-decision-pipeline-for-models-that-score/</link><guid isPermaLink="true">https://www.aimlfieldguide.com/news/2026-11-10-hugging-face-adds-a-decision-pipeline-for-models-that-score/</guid><description>Transformers gains a &quot;decision&quot; task that returns a probability over predefined answers, as a new arXiv paper asks how such models fit into long-horizon agents.</description><pubDate>Tue, 10 Nov 2026 00:00:00 GMT</pubDate><category>transformers-and-llms</category><category>prompting-and-agents</category><category>classical-ml</category><category>huggingface</category><category>opensource</category></item><item><title>Anthropic releases Claude Fable 5.2 and Mythos 5.2 with a paid fast mode</title><link>https://www.aimlfieldguide.com/news/2026-11-09-anthropic-claude-fable-5-2-mythos-5-2/</link><guid isPermaLink="true">https://www.aimlfieldguide.com/news/2026-11-09-anthropic-claude-fable-5-2-mythos-5-2/</guid><description>The two 5.2 models share one underlying network. Anthropic reports faster output, a 2.5 times fast mode at double the price, and about 55 per cent fewer safeguard blocks than Fable 5.1.</description><pubDate>Mon, 09 Nov 2026 00:00:00 GMT</pubDate><category>transformers-and-llms</category><category>safety-security-and-privacy</category><category>evaluation-and-benchmarks</category><category>anthropic</category></item><item><title>Google adds TPU 8tx for confidential compute and the dual-chip TPU 8i Max</title><link>https://www.aimlfieldguide.com/news/2026-11-09-google-tpu-8tx-and-8i-max/</link><guid isPermaLink="true">https://www.aimlfieldguide.com/news/2026-11-09-google-tpu-8tx-and-8i-max/</guid><description>TPU 8tx pairs TPU 8t silicon with AMD EPYC hosts for confidential inference and small-scale training. TPU 8i Max packs two 8i chips into one package. No prices or dates yet.</description><pubDate>Mon, 09 Nov 2026 00:00:00 GMT</pubDate><category>hardware-and-acceleration</category><category>mlops-and-deployment</category><category>safety-security-and-privacy</category><category>google</category><category>pytorch</category></item><item><title>Ollama 0.40.16 ships Ripple-65B Preview, a dense Apache-2.0 coding model</title><link>https://www.aimlfieldguide.com/news/2026-11-09-ollama-0-40-16-ripple-preview/</link><guid isPermaLink="true">https://www.aimlfieldguide.com/news/2026-11-09-ollama-0-40-16-ripple-preview/</guid><description>Three Ollama releases in a week add Ricursive&apos;s 65 B dense model, MiMo V2.6, YaRN long context for Qwen3, and stop OpenAI-compatible endpoints overriding a model&apos;s sampler defaults.</description><pubDate>Mon, 09 Nov 2026 00:00:00 GMT</pubDate><category>open-source-ml</category><category>mlops-and-deployment</category><category>transformers-and-llms</category><category>opensource</category><category>apple</category><category>nvidia</category></item><item><title>Moonshot AI publishes Kimi K3.1 Pro weights, a 2.8 T total, 104 B active multimodal MoE</title><link>https://www.aimlfieldguide.com/news/2026-11-05-moonshot-kimi-k3-1-pro-open-weights/</link><guid isPermaLink="true">https://www.aimlfieldguide.com/news/2026-11-05-moonshot-kimi-k3-1-pro-open-weights/</guid><description>The weights ship in MXFP4 under Moonshot&apos;s own licence, which adds a revenue threshold for hosting providers. Context is 1 M tokens and the card recommends vLLM, SGLang or TokenSpeed.</description><pubDate>Thu, 05 Nov 2026 00:00:00 GMT</pubDate><category>transformers-and-llms</category><category>open-source-ml</category><category>efficient-ml</category><category>opensource</category><category>huggingface</category></item><item><title>llama.cpp v0.9.0 adds OCP FP8 E4M3 quantisation and a batched tensor copy API</title><link>https://www.aimlfieldguide.com/news/2026-11-04-llama-cpp-0-9-0-fp8-e4m3/</link><guid isPermaLink="true">https://www.aimlfieldguide.com/news/2026-11-04-llama-cpp-0-9-0-fp8-e4m3/</guid><description>The November release brings an FP8 file type, a more efficient unified KV cache for multi-sequence serving, four new architectures and ggml v0.29.0 with initial multi-node tensor parallelism.</description><pubDate>Wed, 04 Nov 2026 00:00:00 GMT</pubDate><category>efficient-ml</category><category>open-source-ml</category><category>hardware-and-acceleration</category><category>opensource</category><category>apple</category><category>nvidia</category></item><item><title>Meta buys Conductor, the Mac app that runs Claude Code and Codex agents in parallel</title><link>https://www.aimlfieldguide.com/news/2026-11-04-meta-acquires-conductor/</link><guid isPermaLink="true">https://www.aimlfieldguide.com/news/2026-11-04-meta-acquires-conductor/</guid><description>Terms were not disclosed. Conductor Cloud goes read-only on 13 November and shuts on 20 November, and the local app stops receiving updates in early 2027 while the team joins Meta.</description><pubDate>Wed, 04 Nov 2026 00:00:00 GMT</pubDate><category>prompting-and-agents</category><category>careers-and-industry</category><category>mlops-and-deployment</category><category>meta</category><category>anthropic</category><category>openai</category></item></channel></rss>