Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Device Bound Session Credentials in Asp.net Core(blog.maartenballiauw.be ↗)
    discuss
  2. Embeddable Interactive Avatars(synthesia.io ↗)
    discuss
  3. The paradox of AI automation at the workplace
    discuss
  4. I didn't sign the Fields medallists' letter(gowers.wordpress.com ↗)
    discuss
  5. Artificial Analysis Capability Indices v1.1(artificialanalysis.ai ↗)
    discuss
  6. Show HN: AskMe – answers repetitive support questions(ask-me.studio ↗)
    discuss
  7. Automattic execs signed reciprocal severance deals in Mullenweg absence(techcrunch.com ↗)
    discuss
  8. Ask HN: How do you use AWS AssumeRole?
    discuss
  9. Reckless Expansion Plans Blow Past European Airports Carbon Limits(cleantechnica.com ↗)
    discuss
  10. What Is a Docker Container?(freipaul.com ↗)
    discuss
  11. AI Safety Is Mostly a Sex Cult(skywriter.blue ↗)
    discuss
  12. Stress at work linked to erectile dysfunction, even in younger men(scimex.org ↗)
    1comments
  13. Show HN: Airy TTS – 2K characters/second, $2 per 1M characters(airy.so ↗)
    discuss
  14. Epstein had cache of child sex pics; victims sue to find out who's in them(arstechnica.com ↗)
    discuss
  15. GLM Built Its Own Inference Infrastructure(z.ai ↗)
    discuss
  16. Nobody Encrypts Container Images(bschaatsbergen.com ↗)
    discuss
  17. Jev Is the Missing Piece in Production AI Systems(nimendra.xyz ↗)
    discuss
  18. Release of CmdStan 2.40(mc-stan.org ↗)
    discuss
  19. Backup Exec(wikipedia.org ↗)
    discuss
  20. (Go) Spelunk(github.com/detro ↗)
    1comments
  21. One Year of Sponsored Servo Development(servo.org ↗)
    2comments
  22. EpicSemi Contrail AIX RISC-V Servers(epicsemi.com ↗)
    discuss
  23. Measuring and Exploiting Implicit Trust in LLM Tool-Calling Pipelines(arxiv.org ↗)
    discuss
  24. Axiom Knowledge graph: A deterministic search engine
    2comments
  25. Developer builds 3D code visualizer that takes 21GB RAM shows 2.5M LOC at 120fps(tomshardware.com ↗)
    discuss
  26. CCC invites all model citizens to 40C3(ccc.de ↗)
    discuss
  27. ImageToSTL – photo to 3D, or a logo into a printable STL(imagetostl.im ↗)
    2comments
  28. Code Contracts(code-contracts.cc ↗)
    discuss
  29. Show HN: Gliese Watch – alerts when control over a DeFi contract changes(gliese.io ↗)
    discuss
  30. The Surprising Health Benefits of Reading a Book(time.com ↗)
    discuss

Axiom Knowledge graph: A deterministic search engine

3 pointsby 50m ago
2 comments
Every time you search on internet you might feel in any corner of your mind as does this a correct information because now in 2026 every search becomes AI search and as we know AI can show errors and shows misleading information confidently which is wrong. We want a proper search engine which can show real facts and true facts with under millisecond of search. So, presenting Axiom a search engine which can solve these all problems. Axiom is not a simple search engine but a hybrid BM25 and a sematic vector search engine. AXIOM addresses these systemic pathologies by re locating computational control to a sovereign local exe cution node. AXIOM operates independently of external cloud services, processing web documents through a high-performance pipeline and indexing them into a locally stored vector space. AXIOM is explicitly architected for environments requiring high throughput, structural permanence, and absolute data privacy: • Offline-First Academic Research: Enables continuous, local ingestion of scientific repositories (e.g., ArXiv, PubMed, IEEE Xplore, CERN, Springer) for local query resolution without web access. • Deterministic Agentic RAG: Serves as a zero hallucination ground-truth context provider for autonomous AI agents requiring verified factual lookups. • Self-Hosted Knowledge Infrastructure: Provides high-density indexing of sensitive internal documentation alongside public domain datasets within a strictly air-gapped security perimeter. AXIOM presents a production-grade alternative to centralized search platforms. By unifying high-concurrency Rust spidering, local vector indexing in Qdrant, and strict cross encoder thresholding, AXIOM delivers high throughput information retrieval while ensuring absolute data sovereignty and factual accuracy. Axiom had won from Bing copilot search in information from Wikipedia about maths formulas and concepts and here Bing copilot search told this completely wrong and incorrect whereas on Axiom it shown correct and actual information.