Skip to contentSkip to stories

Updated

#Other

Sep 24

Sep 24Thu
  1. Goodfire ResearchAI score57

    Goodfire finds sparse autoencoder features capture curved neural geometry in three ways

    AIGoodfire Research examines how sparse autoencoder directions relate to curved manifolds in neural representations, identifying shattering, compact capture, and dilution as three ways lines can represent them. The team trained an autoencoder on synthetic data containing shapes such as donuts, spheres, and Möbius strips, and reports that real features in Llama 3.1 8B show dilution. It also describes an unsupervised pipeline that clusters features by firing patterns to surface manifolds in that model.

Sep 23

Sep 23Wed
  1. Engineering at MetaAI score43

    Meta Brings Private Processing to AI Glasses via Confidential Cloud Computing

    AIMeta is extending its Private Processing confidential computing infrastructure to AI glasses, running AI models inside confidential virtual machines so that even Meta cannot access user data. The system relies on hardware Trusted Execution Environments, with remote attestation checked by clients before any data is sent. Meta first introduced Private Processing in 2025 for WhatsApp and the Meta AI app.

Sep 22

Sep 22Tue

Sep 21

Sep 21Mon
  1. Baidu Inc.AI score16

    Kooko, formerly Oreate AI, reaches 10M+ international users in year one

    AIKooko, formerly Oreate AI, has reached more than 10 million international users across nearly 200 countries and regions in its first year. The all-in-one AI workspace can research and analyze data from user-provided files and sources, and turn briefs into documents, presentations, and spreadsheets refinable in a built-in Office editor. It also retains memory of edits, preferences, and completed work for future tasks, and is available on desktop and mobile.

Sep 19

Sep 19Sat

Sep 17

Sep 17Thu
  1. OpenBMBAI score40

    OpenMed and MiniCPM5-2B demo local agentic clinical AI workflow

    AIOpenMed paired with MiniCPM5-2B to demonstrate a local clinical AI workflow combining privacy-preserving data processing with a compact model's tool use and long-context reasoning. OpenMed masks sensitive identifiers and extracts clinical context before MiniCPM5-2B calls tools, compares lab results, and generates clinical handoffs with source references. The post presents this as an example of keeping inference on local, resource-constrained hardware.

Sep 16

Sep 16Wed
  1. WanAI score15

    Ad from Zero contest: create AI ads for fictional products with Wan3.0

    AI架空の商品に、どんな広告をつくる? AZ8 × Wan「Ad from Zero」が開催中です。自由に考えた商品を、Wan3.0で一本のAI広告に。賞金総額3,000米ドル、制作クレジットも用意されています。 応募は9月24日まで。あなたのアイデアを、広告にしてください ↓ Translation: What kind of ad would you make for a fictional product? The AZ8 × Wan "Ad from Zero" contest is now open. Turn a freely imagined product into a single AI ad with Wan3.0. A total prize pool of US$3,000 and production credits are also available. Entries close September 24. Make an ad from your idea ↓

Sep 13

Sep 13Sun

Sep 12

Sep 12Sat
  1. Epoch AI · The Epoch BriefAI score60

    Epoch Brief covers Huawei chips, Nvidia's GDP effect, and GPT-6 Astra benchmarks

    AIEpoch AI's newsletter reports that Huawei is far behind Nvidia and is unlikely to catch up this decade due to export controls. It also finds official US GDP statistics understate growth by about 0.3 percentage points over the past year, and that GPT-6 Astra set new records on Epoch's evaluations, including the Epoch Capabilities Index.

    Why it matters: The newsletter bundles several analyses of AI chips, GDP measurement, and benchmarks, so it helps readers scan the research agenda behind each finding.

Sep 11

Sep 11Fri

Sep 10

Sep 10Thu

Sep 9

Sep 9Wed

Sep 8

Sep 8Tue

Sep 7

Sep 7Mon
  1. Chips and CheeseAI score46

    Arm Unveils C2-Ultra CPU and G2-Ultra NX GPU IP for Flagship Phones

    AIArm's C2-Ultra CPU core claims a 15% peak uplift over C1-Ultra, but Chips and Cheese found the average gain is about 3.2% after accounting for an 8.5% higher clock and a larger 3 MB L2 cache. Arm says the core uses 38% less power, though that figure includes node and implementation changes. Arm calls the Mali G2-Ultra NX its largest GPU re-architecture in seven generations, adding a matrix accelerator that supports INT8 and INT16 but not FP8 or BF16.

Sep 5

Sep 5Sat
  1. AI at MetaAI score38

    AIRA₃ ensemble places 8th with gold-medal results in live competition

    AIMeta's AIRA₃ entered the live competition with an ensemble of models, and the 8th-ranked gold-medal entry combined GPT 5.5 (w/ OpenCode) and Claude 4.8 (w/ ClaudeCode). Post-hoc testing found Muse Spark 1.2 (w/ MuseCode) also reached gold-medal level, while Muse Spark 1.1 (w/ OpenCode) and GLM 5.2 (w/ OpenCode) reached silver-medal level, all graded on the same private test set.

Sep 4

Sep 4Fri

Sep 3

Sep 3Thu