Diagnose Distributed AI Workload Failures with NVIDIA Attribution
In distributed AI workloads, a single error can produce a cascade of symptoms. A GPU or fabric event may surface as NCCL errors on other ranks, scheduler state...
AI NEWS JOURNAL
Search AI news
Search headlines, summaries and articles to connect the dots.
GitHub Copilot Fusion Launches with Dual Capabilities: Intelligent Model Orchestration and Local Model Routing
On October 28, 2026, GitHub released the official version of Copilot Fusion, simultaneously introducing an experimental local model routing feature. This capability enables automatic invocation of local models such as MAI Code 1.1 Flash on Windows devices and activates intelligent routing within VS Code, the CLI, and the Copilot App. The system automatically selects and combines cloud models—like those from OpenAI, Anthropic, and Google—based on three effort levels: Standard, High, and Max, achieving performance close to Claude Code while cutting costs in half during coding evaluations.
A Comparative Study of Parallel Implementations of Short-Time Fourier Transform
arXiv:2610.38069v1 Announce Type: new Abstract: The short-time Fourier transform (STFT) computes the Fourier transform of a short-duration window of a longer signal, which characterizes how a frequency changes over time. While the Fourier transform algorithms, particularly the Fast Fourier Transform (FFT) algorithm, has been well developed to take advantage of parallel computing, there is limited work on understanding the performance of parallel short-time Fourier transform algorithms. In this study, two parallel short-time Fourier transform algorithms, one is FFT-based and the other a quasi-parallel feedforward algorithm, are implemented and analyzed in comparison with each other. It is found that while the quasi-parallel algorithm outperforms in runtime under low concurrency settings, th
FlatCake: A Flat KVCache Storage Architecture for LLM Serving
arXiv:2610.38355v1 Announce Type: new Abstract: KV cache reuse accelerates LLM serving, but existing systems manage host DRAM and NVMe SSD as a tiered hierarchy where DRAM acts as an intermediate cache. Production LLM workloads break both premises of this hierarchy. First, bimodal reuse defeats DRAM temporal filtering, since workloads are dominated either by one-shot blocks with long reuse distances or by multi-gigabyte agentic sessions that exceed DRAM capacity, putting SSD on the critical read path. Second, DRAM staging and eviction-ordered layout bottleneck SSD restoration. Holding host staging buffers for in-flight prefixes causes contention under load that aborts prefetches, while incremental eviction scatters a session's blocks across the SSD and prevents large sequential reads. We p
Scoring Explanations Without Reading Them: Structural Shortcuts in Automated-Interpretability Evaluation Data
arXiv:2610.37676v1 Announce Type: new Abstract: Automated interpretability scores an explanation of a sparse-autoencoder (SAE) feature by asking a model to predict, token by token, where the feature fires on evaluation examples or to judge whether marked tokens fit it. We find that the evaluation data standard pipelines build can be scored without reading the explanation. On Neuronpedia's stored fuzzing runs for 500 random GPT-2 small features, 91.3% of activating examples are cropped with the peak on the centre token, and a rule that says "yes" when the centre token is marked reaches balanced accuracy 0.979, against 0.838 for the stored LLM scorers (recomputed from the stored per-item verdicts, on marks rebuilt from the pipeline's code); a rule reading only the mark count reaches 0.969. T
Wabi reels in $55M for its consumer-focused AI assistant
Wabi Inc., the developer of an artificial intelligence assistant for consumers, today announced that it has raised $55 million in funding. IVP led the seed round with contributions from Amazon Alexa Fund, Andreessen Horowitz and Forerunner. They were joined by a group of angel investors that included Salesforce Inc. CEO Marc Benioff and Shopify Inc. […] The post Wabi reels in $55M for its consumer-focused AI assistant appeared first on SiliconANGLE.
Granola crunches up a dictation feature to compete with Wispr
Uno reverse! After Wispr added meeting notes to its dictation app, Granola is adding dictation to its meeting notetaker. The voice AI wars are getting interesting.
GitHub releases Copilot Fusion, an automated multi-model orchestration tool that boosts coding efficiency and cost-effectiveness.
At Universe 2026, GitHub unveiled Copilot Fusion, an intelligent orchestration layer that automatically routes and combines models from OpenAI, Anthropic, Google, and others to complete coding tasks—eliminating the need for developers to manually select models. It integrates existing automatic model-selection technologies with Project HydraFusion, previewed in September. Once the model selector is enabled, the system automatically adopts execution modes such as single-model, cascading, or critique, based on task complexity.
GitHub Copilot app, Universe releases
Work doesn’t always fit in a single repository or session. A feature can span several repositories. A launch plan can pull in specs, designs, and assets without touching a repository… The post GitHub Copilot app, Universe releases appeared first on The GitHub Blog.
GitHub Copilot Fusion now supports selection of local models
Today we are shipping intelligent routing for local models into experimental preview. GitHub Copilot Fusion can now choose between supported on-device models and cloud models, enabling intelligent local model routing… The post GitHub Copilot Fusion now supports selection of local models appeared first on The GitHub Blog.
NVIDIA RTX Spark Windows Laptops Gain New Models, Agent Tools and Games
NVIDIA RTX Spark powers thin and light laptops built for everything that developers, creators and gamers do. RTX Spark laptops are now available from ASUS, Dell, HP, Lenovo, Microsoft and MSI. These powerful systems combine an NVIDIA Blackwell RTX GPU and an NVIDIA Grace CPU in a single superchip with up to 128GB of unified […]
b11954
tests : adjust constants to reduce irrelevant warnings (#31254) **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11954/llama-b11954-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b11954/llama-b11954-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b11954/llama-b11954-xcframework.zip) **Linux:** - [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11954/llama-b11954-bin-ubuntu-x64.tar.gz) - [Ubuntu arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b1
b11953
cuda : add top-k MoE fusion for 384 experts (#31288) Assisted-by: Codex **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11953/llama-b11953-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b11953/llama-b11953-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b11953/llama-b11953-xcframework.zip) **Linux:** - [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11953/llama-b11953-bin-ubuntu-x64.tar.gz) - [Ubuntu arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/dow
b11951
chat : stop qwen3-coder reasoning and content before template separators (#30557) * chat : strip trailing whitespace from parsed reasoning and content Assisted-by: Claude Fable 5.1 * chat : stop qwen3-coder reasoning and content before template separators Assisted-by: Claude * tests : remove the Qwen parser case from the chat session tests Assisted-by: Claude * common : handle partial prefix until match at end of generation --------- Co-authored-by: Zeshawn Zahid Co-authored-by: Alde Rojas **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11951/llama-b11951-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel
[AINews] Anthropic ends Coding Interviews: "Hunger beats tenure"
Since coding is solved, why code for the job?
Making machines that can listen
Professor Paris Smaragdis develops computers with superhuman hearing for applications like speech recognition, music understanding, and environmental monitoring.
Is Token Confidence Just Chain Length? A Signal-Processing Audit of Reasoning Traces
arXiv:2610.36709v1 Announce Type: new Abstract: Reasoning language models answer math problems by writing long chains of thought, and sampling many chains per problem and taking a majority vote raises accuracy at a large cost in generated text. A popular way to cut that cost scores each chain by the model's confidence in each token it writes, where a token is a word or part of a word. The method averages this confidence over a sliding window of 2048 tokens and keeps or stops chains based on the lowest average. We treat the confidence, recorded token by token, as a signal in the sense of audio or radio, and examine the method with standard signal-processing tools such as spectral analysis and statistical tests for detecting a sudden change. We sampled 16 chains for each of 500 problems from
b11946
server : add realtime conversation and transcription on /v1/realtime (#31269) * server : add realtime conversation and transcription on /v1/realtime Co-authored-by: Xuan Son Nguyen Assisted-by: Claude Opus 5.5 * examples : default the realtime client to the server port 9931 Assisted-by: Claude Opus 5.5 * server : sample realtime responses right after their own decode Slots respond in turns on one context: sampling at the start of the next step read the logits of whichever slot decoded last, so concurrent sessions mixed their outputs. Assisted-by: Claude Opus 5.5 * move more stuff to server-realtime, making it self-contained * examples : fix the type check of the realtime client Assisted-by: Claude Opus 5.5 * examples : use the model card prompts in the realtime client Assisted-by: Claude O
b11945
model : add LiquidAI/d1-2B and LiquidAI/d1-omni-2B (#31275) * model : add LiquidAI/d1-2B and LiquidAI/d1-omni-2B Assisted-by: Claude Opus 5.5 * convert : keep the speech task templates out of the d1 GGUFs Assisted-by: Claude Opus 5.5 * small refactoring: no_batching --------- Co-authored-by: Xuan Son Nguyen **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11945/llama-b11945-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b11945/llama-b11945-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b119
b11944
server: expose the reasoning effort levels of the chat template (#31271) * server: expose the reasoning effort levels of the chat template The capability probe renders a user turn for each candidate level (minimal, low, medium, high, xhigh, max) and for an unknown level. A level is honored when it renders without raising and differently from the unknown level, which shows the template fallback. Levels sharing a render are aliases and the highest one is kept. The caps serialize to JSON, and chat_template_caps in /props reports the list as reasoning_effort_levels, from lowest to highest. * jinja: keep the caps free of the json dependency The caps keep reporting their flags as a map, and the effort levels stay a plain member. common_chat_templates_get_caps() builds the JSON the server reports