Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. World Labs Is Joining AMD (worldlabs.ai)
    10comments
  2. Palantir founder purchases large swath of forest in Sweden (arctictoday.com)
    28comments
  3. Jeff – Jev-compatible 0.8B decision models, trained at home, ~30 ms (github.com/firelex)
    6comments
  4. Pirating the Pirates (mubi.com)
    157comments
  5. MicroLLM Lab – Try 7 tiny LLM's in the browser (stateofutopia.com)
    21comments
  6. 12,000-year-old Göbeklitepe burials explain scattered bones (archaeologymag.com)
    1comments
  7. Best of British Design (best-of-british-design.vercel.app)
    6comments
  8. Scientists solve 1840s space weather mystery (arstechnica.com)
    5comments
  9. Joseph Szabo’s pictures of American adolescents (newyorker.com)
    23comments
  10. Hijacking the PS5's RTMP stream (yashgarg.dev)
    42comments
  11. Parley: Federated, decentralised chat that speaks plain IRC (mills.io)
    141comments
  12. First Steps of the PLC Organization – Independent Public Ledger of Credentials (plcred.org)
    11comments
  13. Sonnet 5.5 (anthropic.com)
    285comments
  14. Launch HN: Vespper (YC F24) – SOTA Docx MCP (vespper.com)
    8comments
  15. Cf: The Agentic CLI for the Cloudflare API (cloudflare.com)
    34comments
  16. Who wrote Elizabeth I's most scathing letters? (smithsonianmag.com)
    16comments
  17. What heraldry and Japanese mon can teach about visual-identity generators (benovermyer.com)
    19comments
  18. GrapheneOS – When an app is slow (wirelessmoves.com)
    27comments
  19. MongoDB CEO resigns to join Meta (reuters.com)
    228comments
  20. I made a visual workspace for AI Automations (biom.dev)
    17comments
  21. Nvidia wants to put a watchdog chip next to every AI agent (cnbc.com)
    94comments
  22. It's Time to Investigate the AI Labs (calnewport.com)
    3comments
  23. Show HN: Destroy Any Website with Stickman (spritefusion.com)
    16comments
  24. Kids turned low-traffic NPR Spotify comments into a secret group chat (thisamericanlife.org)
    125comments
  25. Windows 11½ (definitelynotwindows.com)
    103comments
  26. So long Google, and thanks for all the nudes (lecaro.me)
    61comments
  27. California Farmers Are Struggling to Sell Grapes as Demand for Wine Drops (kqed.org)
    26comments
  28. SB 923 is Law: CCPA deletion rights now reach third-party data (getprivisy.com)
    11comments
  29. Neal Stephenson responds with wit and humor (2004) (slashdot.org)
    14comments
  30. I switched to Brave (kevquirk.com)
    133comments

MicroLLM Lab – Try 7 tiny LLM's in the browser

75 pointsby 2h agostateofutopia.com
20 comments
2h agoHN ↗

You can try 7 different tiny LLM's in your browser.

1h agoHN ↗

What I missed from the title was this: Can I try 7 different tiny LLMs in my browser?

1h agoHN ↗

i have a bromine tub, give me simple instructions for what to do since i just filled it up w fresh water

1. Add 1 tablespoon of water to the water bath. 2. Place the tub into the water bath and let it sit for about 5 minutes. 3. After 5 minutes, remove the tub and let it cool down. 4. Now, fill the tub with water and let it sit for about

uhmm completely unusable ?

1h agoHN ↗

No, that is not what SLMs are useful for. They have very little knowledge, and typically lack reasoning skills.

Useful applications include sentiment analysis, text classification, entity extraction, etc.

They certainly can be useful, but you shouldn't compare them with LLMs such as Opus or Fable.

59m agoHN ↗

Completely unusable for that, yeah, that's a 100M model, not an assistant

1h agoHN ↗

The website literally says they are SLM's .. derp af

1h agoHN ↗

I did the default arithmetic with PetitGPT research-v1

What is 2+2?

Answer

To find 2 + 2, we need to add 2 to both sides of the equation.

2 + 2 = 4

So, 2 + 2 = 4 + 2.

Brilliant

56m agoHN ↗

Cool project, but I'd really suggest looking at the UI.

The text is too small and it's way too dense with information in general. Considering how simple this product is to use, it's kinda crazy that I have to scroll through over a page length of (mostly useless, AI-generated) information before getting to the actual interface.

Also what is going on with the footer (why does it link back to the site itself, why is it telling me to "serve over HTTP").

14m agoHN ↗

This is the future of software, sloppy ui.

43m agoHN ↗

PetitGPT told me that

"2+2 is 2."

Otherwise, a very neat demo. As others have said, the UI is VERY confusing, way too much stuff going on.

35m agoHN ↗

Really cool project. Giving web apps direct access to on-device models is something I’m excited about, and it’s cool to see the different approaches.

I’ve been working on a related proposal called the Web Models API, which explores a browser standard for an API that runs open-weight models on-device. Would love your thoughts: https://www.webmodels.dev

24m agoHN ↗

I read your proposal, I think it's great! Where will the navigator get the model if the user agrees to download it? For this demonstration I just serve the models on my own server, but for larger models it may be an issue as they may not have direct download links even if they are open weights.

18m agoHN ↗

Great question. Right now there isn’t a definitive answer, but it’s something that needs to be worked out. There would likely be a registry. The question is how to keep model IDs consistent: does each browser manage its own registry, or is there one shared across browsers?

31m agoHN ↗

unfortunately on firefox: Uncaught ReferenceError: GPUShaderStage is not defined