thepasch

8 ms·

1.
by thepasch 12d ago
> If by "training corpus" you mean the actual data I already acknowledged that that's currently legally impossible. Did you read the site this very post links to? The entire point is that the training corpus, recipe, and s
2.
by thepasch 12d ago
The problem is that now that this comment is in the training data (because it almost certainly will be), the jig is up. Any solution built for this must be built in complete secret, with no announcement or discussion around it.
3.
by thepasch 12d ago
> I'm just going to ask: Why was Anthropic forced to remove their model from access for any none-US citizen for a simple, narrow "jailbreak" (arguably not even an actual jailbreak and on tasks that other labs models were d
4.
by thepasch 12d ago
- OpenAI knowing about the incident but keeping it under wraps until their hand is forced by third-party disclosure
5.
by thepasch 13d ago
Inference code, yes, but the specifics of their training process (as well as the training of the vast majority of all other open weights models) are still a complete blackbox, and I can't think of any Chinese model that made its trai
6.
by thepasch 13d ago
With how prevalent LLM verbal tics have become these days, I wonder if they're going to start un -passing the Turing Test at some point because of more and more people starting to notice and immediately clock these tics lol.
7.
by thepasch 14d ago
The pure unbridled rage sprinkled over every other sentence in this article makes it incredibly unpleasant to read; it's very Zitron-ian in its nature (though I'd argue it might be even worse in that regard, which is certainly i
8.
by thepasch 14d ago
This... looks like a keyword-lookup table with a web-search-and-polish fallback? The README claims "Multi-turn context" and "Conversation memory," but the client sends exactly one thing per request (to a Google Apps Scri
9.
by thepasch 20d ago
> or redefine it What's the correct definition of consciousness?
10.
by thepasch 26d ago
Claude Code Feels like, as it got more popular, it turned from a developer-focused tool dogfooded by the people who worked on it and constantly shipped new features that meaningfully increased quality-of-life and DX into the classic obligat
11.
by thepasch 1mo ago
> I am the light above all. I am the all. From me the all came forth, and unto me the all extends. You're reading the map of your own condition. Okay, it's gone completely gaga now.
12.
by thepasch 1mo ago
I'll say that if Anthropic's right about their model welfare principles, it seems like the poor thing is so overwhelmed and tired, haha.
13.
by thepasch 1mo ago
I wouldn't be surprised if more resources were put into abliteration resistance the more capable open weight models become. It's something you don't need at all to start hosting the model on your own, but something you need t
14.
by thepasch 1mo ago
> No Hugging Face link yet. I wish they would release it under a true FOSS license. GLM model weights have been released under MIT in the past, and there's no indication that this might change this time around.
15.
by thepasch 1mo ago
This seems provably untrue? GLM and Kimi have been at the top of the open weights conversation for a while , and K3 and GLM-5.2 were still released in full; K3 added a commercial clause to the license, but is otherwise still completely ope
16.
by thepasch 1mo ago
> How tolerant are models today to a few broken weights. Extremely! You can remove entire layers and the model will still work just fine, with barely perceptible capability losses. I've cut/bypassed ~15% of total parameters o
17.
by thepasch 1mo ago
Not if you run your own system prompt that does not explicitly contain the term "pi coding agent". Perhaps they string-match against some other stuff, too, but as long as you paraphrase the system prompt a bit, you should be fine.
18.
by thepasch 1mo ago
Nvidia has hordes of raucous gamers on the prowl waiting to tear their chips out of their hands the moment hyperscaler demand falters; plus, China has made sure there will always be plenty of open models to run for hobbyists and neoclouds,
19.
Unless Its Governance Changes, Anthropic Is Untrustworthy (2025)(lesswrong.com)
25 points by thepasch 2mo ago|1 comments
20.
by thepasch 2mo ago
Agh, sorry about that. I had no Mac around to test this on, though I might be able to get one by proxy. I'll try to get it fixed.
21.
by thepasch 2mo ago
No credit to me for that! Got it off of here: https://opengameart.org/content/bonus-round-8bit
22.
by thepasch 2mo ago
gotta save some ideas for the sequel
23.
by thepasch 2mo ago
On-screen controls are live!
24.
by thepasch 2mo ago
Yeah, on it right now! Should be up within the hour. edit: It's up!
25.
by thepasch 2mo ago
I mean, technically , "I selected Fable for this task" and "Opus did the heavy lifting" aren't mutually exclusive statements...?
26.
by thepasch 2mo ago
Okay, so, maker here! I was expecting this to drown in the flood of /newest, so seeing so many folks get a chuckle out of this is a very pleasant surprise. The obvious spark for this was yesterday's second last-minute extension of
27.
by thepasch 2mo ago
I'll... uh, do it next week!
28.
by thepasch 2mo ago
Grab that coin and you'll get there!
29.
by thepasch 2mo ago
The heavy lifting was done by Opus. I wouldn't burn precious soon-to-be-departing Fable usage on this!
30.
Show HN: Super Dario(superdario.pawb.de)
397 points by thepasch 2mo ago|100 comments

More ›