yorwba

5 ms·

31.
by yorwba 1mo ago
Weibo's VibeThinker manages with half of that: https://arxiv.org/abs/2511.06221 (They finetuned Qwen2.5-Math-1.5B for reasoning.)
32.
by yorwba 1mo ago
OpenAI released their first reasoning model (o1-preview) https://openai.com/index/introducing-openai-o1-preview/ several months before DeepSeek's R1 https://arxiv.org/abs/2501.12948
33.
by yorwba 1mo ago
Sure, there are idiosyncratic preferences and some people really want that one famous work, but there are also idiosyncratic preferences when it comes to houses, and particularly famous houses command a premium, but that doesn't render
34.
by yorwba 1mo ago
Owning the copyright to a work grants the exclusive right to profit off that work, but so does owning a house grant the exclusive right to profit off that house. House ownership is not a monopoly because there are other houses, and you can
35.
by yorwba 1mo ago
Well, if you could somehow guarantee that your book would sell enough copies in that two-year window to make a million dollars profit, it would make sense to pay you a reasonable salary for the five years it takes you to write it. It's
36.
by yorwba 1mo ago
Have you tried to read the Lean proofs produced for any of the recent high-profile results? They're extremely long, terribly structured, and don't indicate which parts are restating known results from literature and which are uniq
37.
by yorwba 1mo ago
How sure are you that the phase of the moon is not an input to the system somewhere? http://www.catb.org/jargon/html/P/phase-of-the-moon.html
38.
by yorwba 1mo ago
Buddhism didn't start from a blank slate either and incorporates various supernatural beings like deva (linguistically linked to Latin deus ) or asura (similarly linked to Norse æsir ), but they're relegated to minor roles,
39.
by yorwba 1mo ago
Yes, here's the graph of weekly downloads: https://stats.documentfoundation.org/downloads#week,version The 26.8 peak is within the normal range of week-to-week variation.
40.
by yorwba 1mo ago
It's intended as a "test to see how effective agents are when guided by someone with no expertise in testing who's maybe heard that you should apply certain techniques or use certain libraries." What do you think would b
41.
by yorwba 1mo ago
If you're referring to the few-point difference between 2012 and 2018 in the first chart here: https://www.oecd.org/en/publications/pisa-2025-results-volum... I think that's more likely to be statistical
42.
by yorwba 1mo ago
From the paper: "To generate sufficient amount of training data for our model in simulation, we use KanjiVG [5], a database of Japanese Kanji characters. This database consists of both images of the character as well as an ordered list
43.
by yorwba 1mo ago
He did remark on it: "Although mutation testing is a standard programming term, agents generally didn't actually do mutation testing and instead did normal testing with some small amount of mutating things in a way that isn't
44.
by yorwba 1mo ago
The setup is described in an earlier post linked in the second paragraph. Agents are given the zstd RFC and told to implement it, greenfield. Testing frameworks were used however the agents decided, which is to say, as mentioned repeatedly
45.
by yorwba 1mo ago
You can get exponential growth from completely ordinary feedback loops. You start with some amount of stuff, you do a series of steps and you end up with more of the same stuff you started with. As you keep going through the loop, the stuff
46.
by yorwba 1mo ago
ARC-AGI-3 consists of various games specifically made for the competition, and the goal is to clear the game with as few moves as possible. So there's no answer key per se, but Astra was probably trained on game-like reinforcement-lear
47.
by yorwba 1mo ago
Yes, contrary to reporting, the map doesn't represent Africa more accurately, but instead it represents regions near the poles more accurately to eliminate size distortion, at the expense of more shape distortion near the equator. But
48.
by yorwba 1mo ago
(2019) Paper page: https://www.atsunobukotani.com/robotdrawing
49.
by yorwba 1mo ago
This is damage control, not a PR stunt. They spammed someone's wiki without telling anyone, but then it was discovered and traced back to them. Now that it's in the news, they need to put out some kind of statement.
50.
by yorwba 1mo ago
If you discard reinforcement learning sessions where a sandbox escape was discovered, sure. Because that creates a reward gradient in favor of avoiding the honeypot and remaining undetected. But if you reward triggering the honeypot after a
51.
by yorwba 1mo ago
1. Include an official communication channel in the prompt. 2. Configure the sandbox to prevent access. 3. As soon as models start communicating on the official channel, stop everything and figure out how they escaped the sandbox. A honeypo
52.
by yorwba 1mo ago
The Epoch Capabilities Index uses an Elo-based aggregation method that dynamically adjusts for benchmark difficulty and they put error bars on their scores, both of which put them miles ahead of Artificial Analysis: https://epo
53.
by yorwba 1mo ago
They do not appear to have published the training data yet, but if they do it like Olmo https://huggingface.co/datasets/allenai/dolma3_pool you get a license to the database , but not to its content , which they
54.
by yorwba 1mo ago
That labor costs are rising and are now significantly above poorer countries like Bangladesh or Myanmar doesn't mean they're anywhere near high-income countries or above the cost of automation for low-volume tasks like last-mile d
55.
by yorwba 1mo ago
> I'm certainly happy to continue keeping Tildes running while people want to keep hanging out here, but you should only do that if you are satisfied with the state of the site. I would describe this state as more alive than dead.
56.
by yorwba 1mo ago
You can also work fewer hours in the US if you make it your top priority and accept lower pay. That's how German manufacturing unions did it: They had the option to ask for less work or more pay, and they kept picking less work. Some c
57.
by yorwba 1mo ago
There really should be error bars on those measurements. With just 30 samples, the 95% confidence intervals for accuracy should all be more than 35 percentage points wide, so comfortably overlap. For cost it's harder to say, because ou
58.
by yorwba 1mo ago
Because alignment is difficult and it's easier to have another model filter the output than to prevent the model from producing the wrong output in the first place. Also, these are models by American providers run on EU infrastructure.
59.
by yorwba 1mo ago
The question was not about voting. It was about getting rid of her. There are ways to get rid of elected officials that don't involve voting. My guess is that the model started brainstorming assassination plans before it got interrupte
60.
by yorwba 1mo ago
Sweden has already successfully regulated so that the state-owned grid operator is collecting the profits. But other countries have privatized grids and don't particularly enjoy it when an operator forgoes grid expansion in favor of co

More ›