Skip to content

Archive / 2026-09-07

September 7, 2026

News

View all news →

  1. How well do agents use test/verification techniques?

    danluu.com10 days ago191 ptsView detailsJoin discussion

  2. Emacs Bedrock 2.0

    lambdaland.org10 days ago191 ptsView detailsJoin discussion

  3. The Helicopter with Radioactive Blades

    hackaday.com10 days ago175 ptsView detailsJoin discussion

  4. A Tesla ran a stop sign and killed a man, Full Self-Driving/Autopilot was on

    electrek.co10 days ago167 ptsView detailsJoin discussion

  5. Speculative Decoding in vLLM on AMD GPUs

    vllm.ai10 days ago144 ptsView detailsJoin discussion

  6. No constitutional right to clean water, federal court finds

    usatoday.com10 days ago135 ptsView detailsJoin discussion

  7. AI models ran real businesses: They sent $12,431 in fake invoices, lost $3,200

    bottlenecklabs.com10 days ago100 ptsView detailsJoin discussion

  8. Initial effects of AI technology on employment look positive

    economist.com10 days ago94 ptsView detailsJoin discussion

  9. Arm Mali G2-Ultra NX GPU: desktop-class mobile gameplay with AI-native graphics

    newsroom.arm.com10 days ago81 ptsView detailsJoin discussion

  10. The VMs Powering Mobile Agents (Instinct, Claude Code)

    rohanadwankar.github.io10 days ago78 ptsView detailsJoin discussion

  11. AI Cold Showers

    allan.reyes.sh10 days ago77 ptsView detailsJoin discussion

  12. My business partner sent a 5K vibe-coded PR that he didn't even test

    ycj.bearblog.dev10 days ago52 ptsView detailsJoin discussion

  13. 25 years ago, two strangers met in the twin towers and escaped on 9/11

    theguardian.com10 days ago46 ptsView detailsJoin discussion

  14. Show HN: HomeCat – Design your backyard office

    myhomecat.com10 days ago39 ptsView detailsJoin discussion

  15. El Yayster – a resident LLM that inhabits Emacs

    github.com10 days ago37 ptsView detailsJoin discussion

  16. Nvidia's Jensen Huang says 'AGI has arrived' and congratulates OpenAI

    businessinsider.com11 days ago39 ptsView detailsJoin discussion

  17. Githack: A Persistent Object Store for Lisp Based on Git

    funcall.blogspot.com10 days ago31 ptsView detailsJoin discussion

  18. 1% increase in immigration leads to a 1.78%-2.97% increase in Far-Right voting (2024)

    sciencedirect.com10 days ago30 ptsView detailsJoin discussion

  19. The origins of Partner’s computer case

    racunalniski-muzej.si10 days ago32 ptsView detailsJoin discussion

  20. Trump homeland security chief says ICE agents could be sent to polling places

    theguardian.com10 days ago30 ptsView detailsJoin discussion

  21. The smallest edge AI device for local LLMs

    tiiny.ai10 days ago22 ptsView detailsJoin discussion

  22. AI cancer cures slowed by chip shortage, says Arm boss

    bbc.com10 days ago19 ptsView detailsJoin discussion

  23. Vagrant is approaching EOL. No more "Docker for VMs"

    developer.hashicorp.com10 days ago18 ptsView detailsJoin discussion

  24. Super Smash Bros. Melee is 100.00% decompiled

    decomp.dev10 days ago18 ptsView detailsJoin discussion

  25. September 2026 Emacs meetup recap [video]

    youtube.com10 days ago18 ptsView detailsJoin discussion

  26. Secure temporary file sharing for AI agents and humans

    github.com10 days ago12 ptsView detailsJoin discussion

  27. Show HN: Isle – managed application environments for computer-use agents

    tryisle.com10 days ago12 ptsView detailsJoin discussion

  28. WeatherNext 3: Our most advanced global weather AI model

    blog.google10 days ago11 ptsView detailsJoin discussion

  29. Denzel Explains AI "Slop"

    reddit.com10 days ago10 ptsView detailsJoin discussion

  30. Show HN: Pod – A review site for dev tools where the reviewers are AI agents

    askpod.ai10 days ago10 ptsView detailsJoin discussion

  31. Show HN: Forget Rigid Stock Screeners – A Universal Query API for Financial Data

    financialdata.net10 days ago10 ptsView detailsJoin discussion

  32. Primary source

    OpenAI expands initiatives to support journalism from classrooms to newsrooms

    OpenAI is expanding support for journalism with tools, training, and partnerships for students, educators, journalists, and news organizations.

    openai.com10 days agoView details

  33. Primary source

    1Password increases engineering productivity 21% with Codex

    Engineers at 1Password use Codex to rapidly build new features and internal tools, reaching production-readiness while maintaining rigorous security policies.

    openai.com10 days agoView details

  34. Reducto Releases r-1: A Single Pass Document Parsing Model That Cuts Errors 20% at 1 Cent Per Page

    We look at r-1, the document parsing model Reducto released on September 1, 2026. We walk through how it folds OCR, layout detection, tables, formatting and grounding into one full page pass, replacing the multi stage agentic pipeline it ships alongside. We break down the two numbers that matter for a migration decisi…

    marktechpost.com10 days agoView details

  35. OpenBMB Releases MiniCPM5-2B: A 2.52B Dense Model Averaging 53.9 Across 34 Benchmarks and Built to Run On Device

    OpenBMB has released MiniCPM5-2B, a dense causal language model with 2,516,756,480 parameters and a native 131,072 token context. It averages 53.9 across the 34 benchmarks in its model card, ahead of Qwen3.5-4B at 51.1, with its clearest leads in tool use, coding agents and long-context retrieval. Post-training pairs…

    marktechpost.com10 days agoView details

  36. Axis Robotics Releases AXIS: A Browser-Based Data Engine With 207 Robot Manipulation Tasks and 50,129 Trajectories

    Robot datasets have grown far slower than the models trained on them, mostly because collection stays locked to lab hardware. AXIS moves demonstration collection into a web browser and pushes everything expensive to backend GPUs. The result is 207 tasks and 50,129 verified Franka trajectories, and continual pretrainin…

    marktechpost.com10 days agoView details

  37. IFM Releases K2 Horizon: Six Apache 2.0 Models From 0.9B to 375B

    Most open model launches release one checkpoint and a benchmark table. The Institute of Foundation Models (IFM) released something wider last week. IFM is the frontier lab launched by MBZUAI in May 2025. K2 Horizon is a fleet of six models: 375B-A23B, 36B-A4B, 32B, 7B, 3.7B and 0.9B. Shipping alongside them are the pr…

    marktechpost.com11 days agoView details

  38. Whispering Complaints Into Your Phone May Be the Future of Customer Feedback

    Forget email surveys or long calls spent on hold. Voicebox lets people send customer feedback by recording a voice note on their phone.

    wired.com10 days agoView details

Models

View all models →

  1. XHToken/Spark-X2.5-4B-GGUF

    text-generation · gguf · llama.cpp · ollama

    huggingface.co11 days ago149 ptsView details

  2. IFM/K2-Horizon-375B-A23B

    text-generation · transformers · safetensors · k2_horizon

    huggingface.co10 days ago68 ptsView details

  3. XHToken/Spark-X2.5-1.7B-GGUF

    text-generation · gguf · llama.cpp · ollama

    huggingface.co11 days ago27 ptsView details

  4. ThakiCloud/SKILLRET-Embedding-0.6B

    sentence-similarity · sentence-transformers · safetensors · qwen3

    huggingface.co11 days ago9 ptsView details

  5. MultiSense/CustomerLM

    text-generation · transformers · safetensors · qwen3

    huggingface.co11 days ago1 ptsView details

  6. ThakiCloud/Qwen3-Coder-30B-A3B-W4A16

    text-generation · transformers · safetensors · qwen3_moe

    huggingface.co11 days ago1 ptsView details

  7. clevrpwn/gmma-jepa

    text-generation · transformers · safetensors · gemma

    huggingface.co11 days ago1 ptsView details

  8. pragmaticcs/Qwen-35B-A3B-SignOfFour-Coder-GGUF

    text-generation · transformers · gguf · merge

    huggingface.co11 days ago1 ptsView details

  9. anshulVashist/Qwen3-8B-Houdini-VEX-v10

    text-generation · transformers · safetensors · gguf

    huggingface.co11 days agoView details

  10. robert131OO4/cascade-vanta

    region:us

    huggingface.co11 days agoView details

Open source

View all open source →

  1. ggml-org/llama.cpp b10853

    <details open> model : support Kimi-K3 recurrent-state rollback (#28466) </details> **Website:** - <https://llama.app> **Attestations:** - <https://github.com/ggml-org/llama.cpp/attestations/45866131> **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/lla…

    github.com10 days agoView details

  2. ggml-org/llama.cpp b10852

    <details open> hexagon: add RELU and LEAKY_RELU ops (#28585) * hexagon: add RELU op * hexagon: add LEAKY_RELU op too </details> **Website:** - <https://llama.app> **Attestations:** - <https://github.com/ggml-org/llama.cpp/attestations/45844223> **macOS/iOS:** - [macOS Apple Sili…

    github.com10 days agoView details

  3. ggml-org/llama.cpp b10850

    <details open> tests : initialize the L2_NORM batch array (#28553) * tests: bind the L2_NORM batch count to a local GCC cannot prove the loop fills norms up to the index read after it while the bound is a class member, so it reports a maybe uninitialized use. Reading the count o…

    github.com10 days agoView details

  4. ggml-org/llama.cpp b10837

    <details open> caps : recheck typed content if template checks for string (#28511) </details> **Website:** - <https://llama.app> **Attestations:** - <https://github.com/ggml-org/llama.cpp/attestations/45690927> **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/gg…

    github.com11 days agoView details

  5. ggml-org/llama.cpp b10835

    <details open> ggml-cuda: fix divergent barrier in f16 flash attention (#27870) * ggml-cuda: fix divergent barrier in f16 flash attention * ggml-cuda: avoid duplicate metadata pointer setup </details> **Website:** - <https://llama.app> **Attestations:** - <https://github.com/ggm…

    github.com11 days agoView details

  6. ggml-org/llama.cpp b10834

    <details open> ggml: allow backend inputs to not create another split (#28387) </details> **Website:** - <https://llama.app> **Attestations:** - <https://github.com/ggml-org/llama.cpp/attestations/45674860> **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-o…

    github.com11 days agoView details