Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. GPT-6 Sol and Luna(openai.com)
    551comments
  2. Claude Opus 5.5(anthropic.com)
    765comments
  3. 'We hacked the FBI:' Hackers say they have data on all FBI employees(404media.co)
    202comments
  4. Microsoft killed FoxPro in 2007. Anyway, here's FoxPro revived(foxscript.org)
    90comments
  5. OpenAI GPT–6 Astra breaks Enigma message that has resisted solution since 2005(cryptocellar.org)
    354comments
  6. ReBarUEFI: Resizable BAR for almost any UEFI system(github.com/xcuri0)
    7comments
  7. SAML: A fractal of bad design(trailofbits.com)
    70comments
  8. Claude Opus 5.5 Intelligence, Performance and Price Analysis (Max)(artificialanalysis.ai)
    62comments
  9. Unreal Agent(unreallabs.ai)
    65comments
  10. WordPress: Unauthenticated path traversal leading to conditional RCE(github.com/wordpress)
    74comments
  11. Native apps written in TypeScript and CSS(github.com/geastack)
    18comments
  12. How did AMD Ryzen get 50% faster in two years?(lemire.me)
    58comments
  13. MUNI Heritage Weekend in San Francisco(lawrence.lu)
    37comments
  14. Show HN: JevBench, a reproducible benchmark for typed decision models(benchmarkheaven.com)
    7comments
  15. OpenAI is well positioned to fast-follow Jev(arcturus-labs.com)
    183comments
  16. Markdown in /src(htmx.org)
    35comments
  17. Show HN: Training a model to identify AI web content from structure alone(arxiv.org)
    8comments
  18. 16-bit Intel 8088 chip (c. 1985)(allpoetry.com)
    13comments
  19. George Lucas Returns to Earth, Bearing Gifts(commonedge.org)
    31comments
  20. The current balance of power in open models(interconnects.ai)
    2comments
  21. The trouble with 'ntile()'(djnavarro.net)
    discuss
  22. The JavaScript Midlife Crisis(maroun-baydoun.com)
    9comments
  23. People hooked on vapes try a new way to quit: cigarettes(bloomberg.com)
    74comments
  24. Apple's Dimensional Drawings(512pixels.net)
    1comments
  25. Launch HN: Coverage Cat (YC S22) – Umbrella insurance via your personal agent(coveragecat.com)
    21comments
  26. Pentagon says overreliance on AI contributed to missile strike on Iran school(bloomberg.com)
    176comments
  27. An update on how we confirm your age group on Discord(discord.com)
    51comments
  28. Apple has added persistent 'ads' to iOS, and it's driving users crazy(techradar.com)
    434comments
  29. There's a high chance of devices being sold with GrapheneOS preinstalled in 2027(grapheneos.social)
    103comments
  30. Did OpenAI solve the wrong Navier-Stokes problem?(scientificamerican.com)
    38comments

LLM Ass Bench

134 pointsby 3h agoassbench.com
37 comments
2h agoHN ↗

The idea that they're going to benchmaxx this in the future amuses me

2h agoHN ↗

It's ok they're wearing thongs it's not nudity.

1h agoHN ↗

still awkward when a page full of thong-covered asses pops up on my extra monitor in my open-seating arrangement office that im required to attend 4 times a week (no of course im not salty or anything)

1h agoHN ↗

You shouldn't ASSume a link with a name like this is sfw.

1h agoHN ↗

its only awkward if you quickly close it as if embarrassed

leave it up all day so others know it's intentional

1h agoHN ↗

jiggle one of the butts while making eye contact with passers-by

1h agoHN ↗

Make an all-team demo on new frontiers in benchmarking technology.

1h agoHN ↗

Have you tried a few toots on an air horn to indicate to your colleagues that the thing on the screen was unexpected?

2h agoHN ↗

Finally, I was getting tired of seeing pelicans on a bike.

1h agoHN ↗

how come opus 5.5 did so well? I'm surprised

1h agoHN ↗

Na, not well at all. Now Fable, fable like'em thick.

1h agoHN ↗

The prompt that created these images doesn't mention that they want a Caucasian skin color. Huh. Racist LLMs?

1h agoHN ↗

Makes sense. Most look like pancake butts or shitty implants.

1h agoHN ↗

It’s only a sample of one per model

1h agoHN ↗

Glad my work blocked this. But can we get a NSFW tag? If this is NSFW?

1h agoHN ↗

If I were you I wouldn't rush to open a website that clearly has "ass" in the title

1h agoHN ↗

i almost wrote a extension to block mentions of pelican bench because i find those so utterly useless. now, Assbench, on the other hand, i could get behind

1h agoHN ↗

"The economy was in a state of deep neglect, a great dust bowl had ravaged food supplies. And the number one movie in the country was called Ass. And that's all it was, for 90 minutes."

1h agoHN ↗

6 months from now we'll be discussing if model training is assmaxxing.

1h agoHN ↗

I'm impressed that Luna got as far as it did on max reasoning. Don't get me wrong, that's one weird looking ass, but even the top GPT model a year ago probably would have done worse based on my experience (with asking it to create meshes in general, not modeling asses... yeah, that's the ticket). Then again, I'm also not, because I've found it to be a great deal at its relatively low price point when the reasoning is set to high.

17m agoHN ↗

Next, HumanBench! Draw an entire, picture-perfect human.