CMay

8 ms·

1.
by CMay 4d ago
You have huge gaps in your understanding of history.
2.
by CMay 8d ago
The reason that many people don't understand how dangerous AI can be, is that listing the real dangers now becomes like a laundry list for less clever people to follow. It's highly unlikely you've ever seen publicly mentione
3.
by CMay 8d ago
The story is both a little overblown and a little underblown, just different parts of it. More people need to be thinking about the feature level of devices and understand that "off" doesn't really mean it most of the time.
4.
by CMay 8d ago
I didn't say a TV had been, if you read. I will absolutely not stop spreading the possibility, because it is in fact a possibility. This is more true now than it has ever been, because even if a manufacturer doesn't do this we&#
5.
by CMay 8d ago
A real problem is that you might not even need to give these internet access. If your neighbor gives their TV internet access and it can establish a mesh connection then it might be able to exfiltrate data that way. I don't know of a
6.
by CMay 8d ago
Does this mean that TVs can never be recycled or resold now, adding to the garbage problem? All you have to do is sell the TV to someone who does eventually connect it to the internet and it will upload all the data it acquired while the p
7.
by CMay 13d ago
This is not a bold claim, this is just how it is. If you ask any average gamer, "if you buy a game on Steam or on a CD from Walmart, do you operate on the understanding that you are now legally allowed to make as many copies of it as y
8.
by CMay 19d ago
> McCarthyism was the pinnacle of the anti-communism movement. How did that turn out, again? It worked out fine? > I'm pretty sure Germany invading other countries was the cause of World War 2. I don't see how you can lump &
9.
by CMay 19d ago
> The solution to save the country is to write off the Constitution in favor of "unitary executive theory" aka autocracy? Unitary Executive Theory is not autocracy. You seem to have capacity to reason and maybe don't unde
10.
by CMay 19d ago
> Every time we impose our imperial will on other countries The US isn't an empire. Is Germany a US territory? Is Japan a US territory? Is Iraq? No. Is Panama? No. Puerto Rico and Hawaii are happily part of the US. We don&#x
11.
by CMay 19d ago
> Lmao you've literally went in there because of your blind anti-communist behaviour after coup-ing your own "allies" of South Vietnam Anti-communism was and still is highly important and critical. Comintern was aiming to
12.
by CMay 19d ago
That wasn't the question you asked. The question you asked was about a couple decades later. When we helped overthrow the previous leader, it was because the fear was that he was establishing communism in Iran. He was unelected, rem
13.
by CMay 19d ago
Yes, after Saddam invaded Iran, because Iran was trying to export its revolution to Iraq. We didn't make Iraq attack Iran though. It was just one more war caused by Iran exporting revolution which has happened several times in the Mi
14.
by CMay 19d ago
We didn't start World War 1, or World War 2, or the Korean war, or the Vietnam war. We didn't make Russia attack Ukraine. We didn't make Iraq attack Iran or Kuwait. We didn't make Iran fund terrorism around the region
15.
by CMay 24d ago
> if you run one model, run glm-5.3 That is a horrible take-away from this, with only 28 tasks and a high pass rate for most models, it says almost nothing. Test a model for your use case and use the fastest, smallest, cheapest model tha
16.
by CMay 28d ago
I think they said they were keeping the old UD 2.0 quant for the larger sizes? So maybe they kept those the same and simply reuploaded them. They said the newer UD 3.0 quant performed worse on some things for the higher quants. So now it
17.
by CMay 1mo ago
Download the original model with the original template, not updated versions of the model or finetuned versions of the model. Then create your own reasoning tests to verify that it is working correctly. You can set a specific seed value to
18.
by CMay 1mo ago
Well in my case I'm using llama.cpp and the mmproj is required, but I think it is just an extracted part of the original model file. Even with audio, yes it technically supports them natively and they're "encoder-free",
19.
by CMay 1mo ago
The updated version will handle tool calling better by default, but the reasoning quality is no longer preserved and is mutilated quite badly.
20.
by CMay 1mo ago
Google provided incorrect settings and an imperfect template. Unsloth modified the template and then finetuned their own version of the model to optimize for some benchmarks as a means of validating quants. Google and Llama.cpp then adopt t
21.
by CMay 1mo ago
For me that moment was Gemma 4 12B QAT. You're not suddenly going to start throwing your hardest programming problems at Gemma 4 12B QAT, it is still 15B parameters less. It's more that, aside from pelican art which isn't w
22.
by CMay 1mo ago
The nature of LLM benchmarking is that they seem to saturate public benchmarks so quick, they are a uniquely efficient case of https://en.wikipedia.org/wiki/Goodhart%27s_law I'm not asking anyone to take my word,
23.
by CMay 1mo ago
> In my experience MTP's speed increase doesn't seem to justify the apparent loss of success at the edge If you set manual MTP settings, you'll override dynamic adjustments the inference engine will try to do. Sometimes t
24.
by CMay 1mo ago
So long as they have flash attention enabled, Llama.cpp enables Sliding Window Attention by default for Gemma 4 models. Even if they're using Ollama or LM Studio I would expect those to mostly be doing the right things.
25.
by CMay 1mo ago
Gemma 4 12B, Gemma 4 12B QAT, Gemma 4 31B, Gemma 4 31B QAT Gemma 4 26BA4B would get close, but not quite and sometimes even get stuck in loops despite a repeat penalty. Do not use any newer updated templates or Unsloth fixes. Use older off
26.
by CMay 1mo ago
Well, I will say: #1: it does not require deep world knowledge, because that's not what local models are for. #2: it directly attacks drive-by understanding, overly linear processing training, poor attention mechanisms, poor reasoning
27.
by CMay 1mo ago
Yeah, Glimmer is excellent. You don't really test Glimmer with one-shots, because it's explicitly designed for multi-turn solution finding. The way I see it, if I've got a task that could be done either agentic or requires
28.
by CMay 1mo ago
Credit where it's due. Qwen 3.8 27B is only the second local model after Gemma 4 that managed to correctly reason through one of my private benchmarks. It took 5x as many tokens to do it and 12m30s with MTP enabled, but it did do it.
29.
by CMay 1mo ago
> For example, there is a legal requirement that Congress authorize military action longer than 60 days. They’re advancing a theory that this is unnecessary or magically resets a timer anytime a cease fire is declared but that doesn’t ma
30.
by CMay 1mo ago
The number of people that agree with something does not inherently make it good or true. History has proven that. Many people don't own guns, partially because violent crime in general has declined significantly over the past 200 year

More ›