big-chungus4

6 ms·

1.
by big-chungus4 6d ago
I think it's that when some code compiles to say assembler, the compiler doesn't prioritize readability and maintainability of the assembler code, since people are not expected to read and maintain it directly
2.
by big-chungus4 7d ago
What are people on about, there is tons to do in NMS, tons of missions, expeditions, build bases, build a corvet, hunt rate parts, ships, rifles, creatures, manage a settlement, automate crafting chains, find a lucrative trading route, and
3.
by big-chungus4 10d ago
That game is great to RL on to teach the model computer use, they might've actually done that for Astra
4.
by big-chungus4 12d ago
Well, do you have any ideas about how to make it scientific? And they clearly needed to do something based on their unnatural ranking
5.
by big-chungus4 16d ago
DeepSeek has been the most used model on open router by far, only temporarily overtaken by ox alpha when it was free, so it's the most representative experience
6.
by big-chungus4 22d ago
They are great base models for fine tuning on both text and visual tasks. Many OCR and object grounding models are based on small Qwen models, though they often replace vision encoder with a bigger one. Qwen3 5-4B is the biggest model I can
7.
by big-chungus4 22d ago
> We are releasing these architectural improvements ahead of time so that the community can prepare for the upcoming full family of Qwen4 models. That gives me hope that "full family" means it will include smaller models like 4
8.
by big-chungus4 22d ago
I hope there is going to be a free endpoint... Unlike 35B-A3B, I am nowhere close to running it locally
9.
by big-chungus4 25d ago
I saw some guy streaming how he was deploying qwen3.8 37B on his local setup. Well, he was asking Claude to do it. It took him two hours of passing errors to Claude for the endpoint to start working, he then started testing it against DS4 F
10.
by big-chungus4 25d ago
MCP can also handle authorization, since you don't want to put your password to skills.md and send it to China
11.
by big-chungus4 1mo ago
So flash is 52 points on artificial analysis, and pro is 53
12.
by big-chungus4 1mo ago
I was recently working on an AI harness too, but I wasn't using AI to code it. It's really easy and requires little code. I wasn't even using langchain - that would require even less code. UI is harder for sure, but it's
13.
by big-chungus4 2mo ago
It was but I'm not sure what in my reply indicates missing it. The comment still made a point and I replied to it
14.
by big-chungus4 2mo ago
None of that requires you to be a good public speaker or speech writer
15.
by big-chungus4 2mo ago
Why exactly is that bad?
16.
by big-chungus4 2mo ago
Qwen3.7 pro is meh, but 3.7 max is a very good model
17.
by big-chungus4 2mo ago
I'd say it's going to be very hard to come up with a method that works on general nonconvex functions while not working on convex functions
18.
by big-chungus4 2mo ago
I'm not sure if the page is not fully loaded for me, because there are a lot of gaps in it, but from what I can see, I still have no idea what "Ode" is. The entire page is just empty marketing without saying anything.
19.
by big-chungus4 2mo ago
Adam doesn't use the second derivatives matrix, it uses second moments of the gradient, which is the diagonal of the uncentered covariance matrix, but neither of them are directly related to SVD or singular values anyway. There is a sl
20.
by big-chungus4 2mo ago
Last statement is a bit sus... Muon computes matrix sign function which can be defined as setting singular values to 1, though you can also define it without SVD. Muon itself doesn't use SVD because it uses a faster method to compute m
21.
by big-chungus4 2mo ago
I am sure they also use Google search and Google maps, and it gives them an even greater advantage
22.
by big-chungus4 2mo ago
A better solution might be not to constrain the generation, but to remove invalid fields from the tool call in the assistant message. So on the next turn, the model receives chat history which contains it's tool call, but without extra
23.
by big-chungus4 3mo ago
You can take the output of the matrix LSTM, which is going to be matrix for each token, and compute the SVD. To get better storage, we want U and V to be the same for all tokens, so that we can operate on the diagonal S matrix. But LSTM is
24.
by big-chungus4 3mo ago
That's useful, but wouldn't help with this particular experiment because they orthogonalize activations, not weights
25.
by big-chungus4 3mo ago
LFM 2.5 1.2B instruct, it's good at instruction following as long as the instruction is short and simple. Gemma 4 E2B is even better than LFM but I don't know if you can fit it in 3GB. Qwen 3.5 0.8B and 2B are smart but very incon
26.
by big-chungus4 3mo ago
I can run Qwen3.6-35B-A3B at 20 TPS on my laptop with RTX 5070 Ti, with partial offloading to RAM. But the most I do is mess with it when I'm bored. I do coding by hand, but I often run autoresearch loops using free models, right now i
27.
by big-chungus4 3mo ago
Just like me! I like AI because of how smart I am.
28.
by big-chungus4 3mo ago
I am interested to know why people use those graph based agentic frameworks. Why not just define the behavior in python?
29.
by big-chungus4 3mo ago
Or attach an LSP server with a type checker and require the model to produce strict strongly typed python code
30.
by big-chungus4 4mo ago
> we would never conclude that the LLM has conjured up digital re-creations of Julius Caesar and Genghis Khan At some point the AI might become so powerful that whatever it reasons through isn't any less real than a computer simulat

More ›