Hacker News: anana_

New comment by anana_ in "Qwen3.8 27B scores 52 on Artificial Analysis"

anana_ — Mon, 17 Aug 2026 18:48:10 +0000

Agreed. Luckily, this model also scores high in AA non-hallucination, so it knows what it doesn't know -- perfect for situations where it can just tool call a web search.

New comment by anana_ in "Qwen3.8 27B scores 52 on Artificial Analysis"

anana_ — Mon, 17 Aug 2026 18:22:54 +0000

And to read the tea leaves a little:

3.8 actually performs slightly worse than 3.6 on AA-Omniscience Accuracy, which could imply that they traded out world knowledge for capability in other areas.

It also produces nearly twice as many tokens per task as 3.6 (and by extension, time), which may be a tradeoff required to achieve correctness at this parameter size.

New comment by anana_ in "Qwen3.8 27B scores 52 on Artificial Analysis"

anana_ — Mon, 17 Aug 2026 17:30:43 +0000

For more context, this puts it on par with models like GLM 5.2 and GPT 5.6 Luna, which are far larger

Qwen3.8 27B scores 52 on Artificial Analysis

anana_ — Mon, 17 Aug 2026 17:25:17 +0000

Article URL: https://artificialanalysis.ai/models/qwen3-8-27b

Comments URL: https://news.ycombinator.com/item?id=49334544

Points: 350

# Comments: 157

New comment by anana_ in "Qwen 3.8 27B"

anana_ — Fri, 14 Aug 2026 21:51:08 +0000

When stuff like this: https://doublespeed.ai/ exists I don't find that hard to believe at all, although it cuts both ways

New comment by anana_ in "Qwen 3.8 27B"

anana_ — Fri, 14 Aug 2026 16:20:41 +0000

Seems like MTP is available immediately!

New comment by anana_ in "Qwen3.8-27B"

anana_ — Fri, 14 Aug 2026 16:19:11 +0000

As was the case with GLM 5.3, it seems that there is still much juice to be squeezed from post-training

New comment by anana_ in "Qwen 3.8 27B"

anana_ — Fri, 14 Aug 2026 15:15:48 +0000

Monstrous benchmarks! Hoping it is not benchmaxxed.

New comment by anana_ in "GLM-5.3: Frontier coding with emergent cyber capabilities"

anana_ — Fri, 14 Aug 2026 05:38:59 +0000

What a week for AI model releases

New comment by anana_ in "Qwen 3.8"

anana_ — Sun, 19 Jul 2026 17:38:30 +0000

Apparently agentic performance in Gemma was improved recently: https://x.com/googlegemma/status/2077449152062247219

Too little too late imo

New comment by anana_ in "Ornith-1.0: self-improving open-source models for agentic coding"

anana_ — Mon, 29 Jun 2026 20:12:49 +0000

They keep mentioning a 31B dense model, but there are no benchmarks or weights for it anywhere?

New comment by anana_ in "Qwen-AgentWorld: Language World Models for General Agents"

anana_ — Wed, 24 Jun 2026 16:58:26 +0000

It looks like the purpose of this model is to i. generate environmental sim data for doing RL on other models or ii. act as a foundation model (they trained it to select actions as well as predicting the next state in the same loop?)

Either way, neither are intended for end consumers.

New comment by anana_ in "Qwen-AgentWorld: Language World Models for General Agents"

anana_ — Wed, 24 Jun 2026 07:42:55 +0000

I believe the benchmark listed is about simulating the environment for the various tasks, rather than doing them. It seems that the point of this model is to generate sim data to improve other models with

New comment by anana_ in "Ask HN: Has anyone replaced Claude/GPT with a local model for daily coding?"

anana_ — Mon, 15 Jun 2026 18:40:06 +0000

Unfortunately on Strix Halo or any similar unified memory set up, dense models are gonna be dirt slow due to the tiny memory bandwidth... But I agree, 27B is superior.

New comment by anana_ in "Ask HN: Has anyone replaced Claude/GPT with a local model for daily coding?"

anana_ — Mon, 15 Jun 2026 18:38:06 +0000

Perhaps try a different model? Just from anecdotal experience, I find that the Gemma models smaller than 31B do not tool call as often as they should.

Some of the benchmarks appear to back this up [0]

Of course, a lot depends how you are using it (inference parameters, harness, prompting, etc.), but the model is quite important too.

[0]: https://artificialanalysis.ai/models/open-source/small?model...

New comment by anana_ in "CrankGPT"

anana_ — Mon, 15 Jun 2026 18:09:43 +0000

I have one too and it never occurred to me to use it for anything other than games. Would be interested in seeing how you did it!

New comment by anana_ in "LinkedIn is scanning browser extensions"

anana_ — Thu, 30 Apr 2026 22:22:24 +0000

They do now - https://support.mozilla.org/en-US/kb/use-sidebar-access-tool...

New comment by anana_ in "Kimi K2.6: Advancing open-source coding"

anana_ — Mon, 20 Apr 2026 18:35:06 +0000

Hypothesizing here, but maybe the idea is sort of a form of technological/economic warfare? Releasing performance equivalent yet more cost efficient open weight models should in theory drive the cost of inference down everywhere.

This I assume will make it more difficult for US AI labs to turn a profit, which might make investors question their sky high valuations.

Any sort of melt down in the AI sector would almost certainly spread to the wider US market.

In contrast, in China, most of the funding for AI is coming directly from the government, so it's unlikely the same capital flight scenario would happen.

New comment by anana_ in "I ran Gemma 4 as a local model in Codex CLI"

anana_ — Mon, 13 Apr 2026 11:27:26 +0000

I'm not saying it's the latest Qwen iteration - that would be Qwen3.6.

I'm saying it's the latest iteration of the finetuned model mentioned in the parent comment.

I'm also not suggesting that it's "the latest and greatest" anything. In fact, I think it's rather clear that I'm suggesting the opposite? As in - how can a small fine tune produce better results than a frontier lab's work?

New comment by anana_ in "I ran Gemma 4 as a local model in Codex CLI"

anana_ — Mon, 13 Apr 2026 09:38:45 +0000

It's rather surprising that a solo dev can squeeze more performance out of a model with rather humble resources vs a frontier lab. I'm skeptical of claims that such a fine-tuned model is "better" -- maybe on certain benchmarks, but overall?

FYI the latest iteration of that finetune is here: https://huggingface.co/Jackrong/Qwopus3.5-27B-v3