XCSme

31 ms·

361.
by XCSme 7mo ago
Good work, it's quite close to Gemini 3 Pro in my tests, but 10x cheaper: https://aibenchy.com/compare/google-gemma-4-31b-it-medium/go...
362.
by XCSme 7mo ago
It's 8.3 vs 8.1, I wouldn't call that significantly better. I think GLM got a bit in front, because on some tests that both got wrong, GLM did sometimes (inconsistently) respond with the correct answer. That being said, yes, in th
363.
by XCSme 7mo ago
The questions do ask specifically to respond with the answer only, with an example format given in many cases. Note that all reasoning models are tested with "medium" reasoning. The benchmarks are questions/data processing ta
364.
by XCSme 7mo ago
Why not? I described this in more detail in other comments. Even when using structured output, sometimes you want to define how the data should be displayed or formatted, especially for cases like chat bots, article writing, tool usage, cal
365.
by XCSme 7mo ago
The main reason is that Claude models tend to ignore instructions. There is a failure example on the Methodology page.
366.
by XCSme 7mo ago
Oh, I didn't think about this, that's a good idea. I also feel generally model performance changes over time (usually it gets worse). The problem with doing this is cost. Constsntly testing a lot of models on a large dataset can g
367.
by XCSme 7mo ago
I used qwen 3.5 plus in production, it was really good at instruction following and tool calling.
368.
by XCSme 7mo ago
Yup, they do quite poorly on random non-coding tasks: https://aibenchy.com/compare/minimax-minimax-m2-7-medium/moo...
369.
by XCSme 7mo ago
But I have to send the same sort of information (albeit shorter) via email on a regular basis. A lot of alerts, reporting, quotes, code snippets, short documentation or step by step instructions, etc. I don't just send emails to say &q
370.
by XCSme 7mo ago
Why isn't this website plain text then?
371.
by XCSme 7mo ago
Lite >For small brands wanting to get started with monitoring and content. >$249/month Is $249/month something most small brands/shops can afford? Many have a few $ks in total revenue.
372.
by XCSme 7mo ago
I think the OP app is meant for creating transactional emails (or bulk-send emails like newsletters). Those templates should account for all types of people and accessibility levels (including things like ADHD, where you need a big red butt
373.
by XCSme 7mo ago
But they are a lot easier to see and click (accessibility, larger hit area). You could have a larger text instead of a button, but changing font size is also HTML and not plain-text anymore.
374.
by XCSme 7mo ago
What about images, links? Formatted text like bold or underline? I also prefer plain text, but in most of my emails I talk about technical stuff, or I send transactional emails that require actions, in which case showing buttons is a much b
375.
by XCSme 7mo ago
Yes, but not necessarily new knowledge. It means extending/expanding something, but the information is based on the current data. In computer games, extrapolation is finding the future position of an object based on the current positio
376.
by XCSme 7mo ago
That was my point: "I am not necessarily saying humans do something different".
377.
by XCSme 7mo ago
So, just connect LLMs to lava lamps? Jokes aside, imagine you give LLMs access to real-time, world-wide satellite imagery and just tell it to discover new patrerns/phenomens and corrrlations in the world.
378.
by XCSme 7mo ago
Change is always hard, even if it will be good in 20 years, the transitions are always tough.
379.
by XCSme 7mo ago
Not sure if AI can have clever or new ideas, it still seems to be it combines existing knowledge and executes algoritms. I am not necessarily saying humans do something different either, but I have yet to see a novel solution from an AI tha
380.
by XCSme 7mo ago
Shameless plug: if you think about switching from Google Analytics/Hotjar, check out my project[0] (built in EU, started in Romania, now working on it remotely in Netherlands). If not, happy to hear any criticism or the alternatives yo
381.
by XCSme 7mo ago
How comfortable are you guys with the fact the US has just partnered with OpenAI to enable mass surveillance?
382.
by XCSme 7mo ago
I've been building my project[0] for 14+ years, but I'm still looking forward to the day I won't have to worry about next month's rent. Customers are happy though. Last week I got messages from two customers that purchas
383.
by XCSme 7mo ago
I tried it, it was fun to get started, but after 3-4 sessions, when you want to start getting good and learning stuff, it's not fun anymore. I tried playing with Playtomic, with better players, and everyone expects to play the same, go
384.
by XCSme 7mo ago
So, we can move the taskbar, more AI, and some paint-flashing fixes
385.
by XCSme 7mo ago
Which year was that?
386.
by XCSme 7mo ago
Hard to tell, they only mention a few ones that got better, not clear results on others
387.
by XCSme 7mo ago
But if it got worse on other tests, it doesn't do much good, right?
388.
Ask HN: Are MiniMax Models Scams?
3 points by XCSme 7mo ago|2 comments
389.
by XCSme 7mo ago
It will be Gemini World instead ...
390.
by XCSme 7mo ago
I have a Quest 2, I only use it to play Eleven Table Tennis VR. I don't understand what Horizon Worlds is or the other thing mentioned.

More ›