Thread

1. JCharante · 1785957691 · focus · HN ↗

   I have done my own testing and found that smaller models can beat their larger siblings on fact retrieval from documents. I haven’t investigated it in depth with a large enough dataset but my guess is that larger models overthink it while smaller ones just do it. I would like if they compared this with 5.6 Luna instead.

   1. andrenotgiant · 1785958098 · focus · HN ↗

      Any data or public links you can share? That surprises me
   2. barake · 1785958146 · focus · HN ↗

      Anecdotally, it feels like Opus, Fable, and Sol "get distracted" when you use them for writing code. Great at reasoning and coordination but they will go off on a tangent and refactor half the code base. I only use them for reasoning (of course) and coordinating subagents.

      1. CoolCold · 1785997669 · focus · HN ↗

         mind sharing hints/links on your harness/flow setup?

         I did several attempts with naive prompting, but spent more time babysitting than actual flow

   3. jorl17 · 1785965463 · focus · HN ↗

      Have been feeling the same. There's a sweet spot that threads the needle between "too dumb to search the right thing / relay the correct results" and "too smart to just stop overthinking and just report the damn thing"
   4. hankbond · 1785968894 · focus · HN ↗

      Just an anecdote but thats why Deepseek v4 flash 0731 is my current favorite model. It's really not very "eager" and stays on the task at hand.
   5. seahyinghang8 · 1785982147 · focus · HN ↗

      we actually have the test benchmark against luna too! it's just not in our title but you can see it in the first diagram below the title. luna does pretty well tbh but sol is just a tad bit better. but luna is way cheaper.

      if you want to dive down into the various traces of the benchmark, you can check this out: <a href="https:&#x2F;&#x2F;app.castform.com&#x2F;train&#x2F;a7a898f6-d802-4908-b044-acb812f14a48?tab=comp" rel="nofollow">https:&#x2F;&#x2F;app.castform.com&#x2F;train&#x2F;a7a898f6-d802-4908-b044-acb81...

      - founder of castform

   6. seahyinghang8 · 1785982383 · focus · HN ↗

      founder of castform here again - slightly unrelated to retrieval but on the topic that folks are discussing here, i was actually collecting benchmarking various coding traces for the purpose of training a model router and surprisingly, luna held up very well against sol and terra. it was able to solve close to &gt;95% the that sol can handle at a fraction of the cost. have not benchmarked the OSS models yet but will add the popular ones to the list like Deepseek Flash and Kimi k3 to see how they fare.

      will share the full results soon!

Unofficial Hacker News client; not affiliated with Y Combinator.