At 16K features, flat autoencoders break. Curved space doesn't

In plain words

HyperSAE is a mechanistic interpretability engine that extracts hierarchical concept ontologies from large language models using hyperbolic sparse autoencoders. It combines the speed of standard Euclidean autoencoders with the semantic mapping capabilities of hyperbolic geometry, addressing limitations of flat autoencoders when dealing with 16K or more features. The tool is designed for researchers and practitioners working to understand and interpret deep learning models.

written from the facts on this page · September 2026

• 
  Llama 3.2 Interpretability with Sparse Autoencoders2024 · github.com · ▲579

  I spent a lot of time and money on this rather big side project of mine that attempts to replicate the mechanistic interpretability research on proprietary LLMs that was quite popular this year and produced great research papers by Anthropic [1], OpenAI [2] and Deepmind [3]. I am quite proud of this project and since I consider myself the target audience for HackerNews did I think that maybe some of you would appreciate this open research replication as well. Happy to answer any questions or face any feedback. Cheers [1]…

• L3
  Llama 3.3 70B Sparse Autoencoders with API access2024 · goodfire.ai · ▲201
• 
  Google EmbeddingGemma 216d ago · blog.google · ▲318

  An open, lightweight multimodal embedding model

• 
  DeepSeek-V3.2-Exp2025 · huggingface.co · ▲155

  Long-context efficiency with DeepSeek Sparse Attention

• 
  Model2vec – Lightning-fast Static Embeddings for RAG/Semantic Search2024 · github.com · ▲28

  We’ve recently open-sourced Model2vec, a method to distill sentence transformers into static embeddings that outperform all previous approaches by a large margin on MTEB. Our new models set a new state-of-the-art for static embeddings. Main features: - Our best model (potion-base-8M) has only 8M parameters, which is ~30mb on disk - Inference is ~500x faster than the distilled base model (bge-base), on a CPU - New models can be distilled in 30 seconds on a CPU without requiring a dataset - just a vocabulary - Numpy-only inference: The packaged can be install the package with minimal…

• 
  Xcessiv – Fully managed web application for automated machine learning2017 · github.com · ▲108

  A web-based application for quick, scalable, and automated hyperparameter tuning and stacked ensembling in Python. - reiinakano/xcessiv

• 
  SimCity, but with Permits▲930

  A free city-building game in your browser. You are the mayor. You know what the city needs. Now get it past the zoning board, the neighbors, the appeals court and the next election.

  Life & fun · 2d ago · permit.petercai.com

• IV
  I've been paying for a rural Tanzanian's education for 10 years▲778

  Life & fun · 14d ago · tanzaniaeducationproject.org

• 
  Bigwords.page – Turn any screen into a sign. The URL is the app▲728

  Turn a phone, tablet or TV into a sign. Type a message and it fills the screen: welcome signs, timers, Wi-Fi passwords and more. No app, no account, free.

  Life & fun · 15d ago · bigwords.page

• 
  Guitoro, a virtual guitar you strum with your keyboard▲573

  Play a free virtual guitar online with your keyboard or by touch. Slow songs down and see every chord, string and strum. Acoustic, classic, electric and bass.

  Life & fun · 9d ago · guitoro.com

• 
  See *almost* any carriers phone settings▲432

  Carrier settings from iPhone, Pixel and Galaxy firmware, decoded and compared: APNs, VoLTE, 5G and Wi-Fi Calling per carrier, and what each build changed.

  Life & fun · 13d ago · carrierexplode.com

• 
  The whole Solar System and every tracked satellite, live in the browser▲407

  Life & fun · 23d ago · space.bl2.net

• TL
  The load-bearing vocabulary of Claude▲710

  Life & fun · Aug 2026 · louisabraham.github.io

• 
  Hey Noah▲656

  A proactive AI executive assistant for founders

  AI · Aug 2026 · heynoah.io

• 
  Clipto MCP▲655

  Hey HN, We recently built Clipto. It’s a tool that lets you search over terabytes of video, audio, and images on your computer, without relying on the cloud. Motivation: we probably all had this similar experience, we know a moment exists in a video or audio, but finding it takes hours scrubbing the timeline. You can send all the media to process in the cloud, but it’s slow, expensive and raises privacy concerns. So we decided to build our own on-device media search engine. How it works (high level): 1. We ingest video, audio, image; normalize formats via ffmpeg; run content analysis to…

  Work · Aug 2026 · clipto.com

• 
  Simple algorithm and color space to generate diverse skin tones▲627

  Hello HN! I found that picking out plausible but diverse skin tones for my digital art and game development projects was kind of difficult, and I got curious about if there was a way to define a color space that made it easy. I've built a color picker and procedural generation algorithm based on the space as well as a bunch of other fun js features and demos throughout the page that use the equations. If you find it interesting, I have lots of explanations of how I built it and what properties the space has. The methodology might be a bit shaky, but hopefully the result is as helpful for…

  Life & fun · Aug 2026 · toneyalexander.github.io

• 
  AdAnt AI▲620

  Claude for viral, high-converting social ads

  AI · Aug 2026 · adant.ai

• 
  Astute▲606

  Automate your B2B brand going viral, with new media creators

  AI · Aug 2026 · company-app.joinastute.com