#Tutorial/How-to

AITencent Hunyuan says it shrank the 770B-parameter Hy4 preview from roughly 1.5TB to 214 GiB while keeping the parameter count unchanged. The quoted Zhihu post by a Tencent Hunyuan quantization team member describes the method: a 1.25-bit sparse ternary encoding, mixed precision across expert layers, and STQ1_0 CUDA kernels in llama.cpp. The author reports nearly unchanged MRCR retrieval and a small decline in math.

Why it matters: The quoted Zhihu post explains how Hy4 preview's weights were quantized and kept usable at inference, a concrete engineering case for compressing large MoE models.

AISpaceXAI says its combined support team handled a 175% rise in tickets without hiring, crediting Grok Bot, which it says would otherwise have required about 200 additional staff. The company reports resolving tickets for $0.20 to $0.30 each, versus the $1 to $4 per resolution it attributes to traditional AI support tools. Grok Bot is also reported to resolve 99% of refund requests without human intervention.

AITogether AI's canary rollouts shift production traffic between two model deployments on the same endpoint in staged percentages, with optional metric gates between steps. Operators can choose canary, blue-green, or rolling strategies, and a rollout starts only when explicitly launched; it can be paused, canceled, or reversed. The platform scales the target before moving traffic and waits for routing to converge before draining the source.

AIMike Knoop notes that logprobs used to be widely exposed by LLM inference APIs and sees the market maturing so that parts of the LLM stack can be packaged in new, useful ways. He links this to Bryan Helmig's post on prompting with max_tokens: 1 plus logprobs for fast, parallel judgments, which Helmig says has a lot more depth than he expected.

AIThe article explains how Mixture of Experts models change inference by making prefill, midfill, decode attention, and decode experts distinct workloads. It describes how KV cache state, expert routing, and parallelism choices shape compute, memory, and network demands across an inference cluster.

AIMax Woolf reports that prompting agents to iteratively optimize code can make it faster than current state-of-the-art libraries. He says the blog post includes his prompts and benchmark results.

AILMSYS Org shared a blog post about NVFP4 KV cache, a topic linked from its 2026-09-16 article. The post itself contains only a link, so no further technical details, figures, or results can be confirmed from this source.

AIA research-agent design has GPT-6 orchestrating while MiniCPM5-2B runs locally on DGX Spark to read sources and extract traceable evidence. The post presents this as an example of pairing frontier reasoning with compact, efficient models for transparent research workflows.

AIMatei Zaharia, a prominent AI researcher, said it is very cool that GEPA works on Jev. The post is a short endorsement, linking to background about a test in which GEPA optimized Jev's prompts for extracting suspected adverse drug effects from medical sentences.

AIFireworks AI reports that an oracle router choosing among 18 models per DeepSWE task reaches 97.6% at $1.88 per task, versus 74.1% at $6.52 for GPT-6 Astra alone. The oracle is hindsight-based, so the authors say a production router must predict the best model before the task starts, which is the harder problem.

AIvLLM achieved 5000 total token throughput per GPU in high-throughput PD serving of Qwen3.8-2.4T on a GB300 NVL72 cluster under an 8K/1K workload. The low-latency scenario reached 180 generated tokens per user, with both results shown on the Pareto frontier. The post also provides srt-slurm recipes and explains the tuning process used to create them.

AIA commenter praises a practical setup where a small local model runs a log-debugging pipeline entirely on CPU through OpenClaw and Ollama. They argue that workflow design can make a small model far more useful than its raw size suggests.

AIOpenBMB's 2B-parameter MiniCPM5 model runs as a local news desk on an older i5-9400F PC with 16GB RAM and no cloud API. The developer built a system that collects official sources hourly and sends a 24-hour Telegram recap with a lead story and links.

AIJasper Lu's new blog post walks through training a search agent with GRPO, showing how small reward function changes teach a model to avoid sloppy tool calls, prune unnecessary documents, and balance persistence against token efficiency. The post makes every rollout browsable and releases the code as open source, with the full process from learning rate sweeps to reward shaping documented.

AIGoogle's Envisioning Studio, with Google Labs, co-developed custom Google Flow tools with designers Jane Wade and Sergio Hudson ahead of New York Fashion Week. Wade's Styling Suite let her style runway looks on digital models before producing physical samples, while Hudson's Runway Visualization helped him stage his show within a tight budget. The source says the tools are built with natural language and no coding experience.

AIHamel Husain and Shreya Shankar's FAQ explains AI evals as tests of whether an AI system does what users and the business want. It recommends starting with error analysis on at least 30 traces, then turning recurring failures into binary code-based checks or LLM judges validated against human labels.

AIA Mariposa Museum exhibit photographed by writer Steven Johnson, shared by Gemini Notebook, documents the Sierra Nevada town in 1859, ten years after the Gold Rush began. Johnson says he used the Gemini Notebook mobile app's camera feature to generate a detailed report from photos of the display, which he says was 99% accurate on fact-checking.

AIJetBrains describes building Air Context, a RAG pipeline that gives LLM agents semantic code search over real repositories instead of grep. The first installment covers parsing, chunking, and vectorization, arguing that fixed-size line chunks split related code and that structure-aware chunking using language grammar produces better retrieval units.

AIZ.ai reports that GLM-5.3 helped build and optimize the inference infrastructure for GLM-5.3-Flash. The system went from first successful run to production readiness in under two weeks, with end-to-end throughput tripling over the initial baseline. The team credited dense feedback from local correctness tests, execution traces, microbenchmarks, and end-to-end measurements for enabling targeted hypothesis testing.

AISundial fine-tuned Thinking Machines' Inkling-Small with RLVR on 3,978 verified TeX.StackExchange fixes, using rewards for compilation and PDF match and penalties for removed content. The trained model fixes 83.7% of LaTeX errors in under one second at $0.0013 per fix, according to the post. Sundial says it is rolling out the model in its editor, applying fixes as suggestions and rebuilding the PDF.

AIMosaic, Ponder Studio, and Revyl used early access to Google's Gemini Flash models to test agentic video understanding on long footage. Mosaic reports a 97% cut in median token usage and nearly double the ability to handle complex edits, while Ponder Studio reports a 0.967 F1 score and about 72% lower token costs for B-roll selection. Revyl says the approach improved mobile UI bug-catching accuracy by 65%. The capability is available now for video uploads and YouTube videos via the Gemini API.

AITogether AI says teams can migrate from closed to open-source models in weeks to months, not years, using a managed service. The guide recommends defining the use case, shortlisting candidates with relevant benchmarks, and replaying existing traffic against them to test accuracy, cost, and latency against current targets.

AIGoogle's Gemini Notebook now offers live spoken Q&A over class materials in about 100 languages, and lets students record lectures on the go with audio notes saved automatically to a chosen notebook. University students in 140+ countries can also still get a free Google AI Plan for higher limits and access to more Google products.

AIGoogle reports that its technologies and products now power everyday interactions in more than 300 languages used by over 7 billion people, about 86% of the global population. The post describes new speech models, including Gemini 3.5 Live Translate and Gemini 3.5 Transcribe, plus the TranslateGemma open translation models trained across 55 languages.

AIPart 2 of the zero-trust agents series moves security checks from agent code to the Gemini Enterprise Agent Platform runtime. Model Armor screens prompts and responses, Semantic Governance Policies judge proposed tool calls against intent and business rules, and Agent Anomaly Detection flags multi-turn drainage that single-turn checks miss. The same Customer Support and Returns Agent from Part 1 is used, with the companion demo open-sourced on GitHub.

Why it matters: The post walks through a concrete refund agent under four attacks, showing how screening, intent judgment, and anomaly detection each catch what the others miss.

AIThe vLLM team trained a DSpark speculative decoding draft model for Kimi K3, a 2.8T-parameter model, using the Speculators library on GB300 NVL72 hardware. They added a MooncakeHiddenStatesConnector to stream hidden states from disaggregated vLLM inference nodes to training nodes across multiple machines. The released speculator raises single-stream interactivity from about 110 to about 435 tokens per second per user on math reasoning, with up to about 3.5x higher output throughput under concurrent load.

Why it matters: The post shows how hidden-state extraction and Mooncake transfers let a 2.8T-parameter model's speculator be trained across multiple nodes, a reusable pattern for similar setups.

AIA hands-on guide by @pandemicsyn walks through writing evals that detect when an AI agent claims to have completed a task it never did. Working through a demo agent that fails on purpose, the author refines the checks until they can distinguish real work from mere claims of work. The post includes a coding agent skill that can guide readers through the exercise.