krackers

12 ms·

241.
by krackers 6mo ago
I don't think fil-c is a drop in C replacement, there are things you can do in C such as certain types of pointer abuse that fil-c prohibits (e.g. cast pointer to int, then back). It's probably easier to port an existing C projec
242.
by krackers 6mo ago
There's some irony there in that the whole maps fiasco lead to firing of Forstall which allowed Ive to become head of design, which basically led to the current state of macOS design. I do wish that some day someone will tell the story
243.
by krackers 6mo ago
Hm interesting, this seems to examine Anthropic's prior work. If I understood the paper right, during post-training (e.g. DPO) models learn the correct "shape" of responses. But unlike SFT they're also penalized for goin
244.
by krackers 6mo ago
Masked language modeling has been compared loosely to text diffusion [1], so the paper's title claim may be loosely true in some sense even if it's misleading. [1] https://nathan.rs/posts/roberta-diffusion&#x2
245.
by krackers 6mo ago
Maybe for very high-level intuition, it's vaguely similar to other randomized algorithms that you want to minimize the worst-case on expectation and the easiest way to do so is to just introduce randomness (think quicksort, which is N^
246.
by krackers 6mo ago
If I have a conversation with claude then come back 30 minutes later to resume the conversation, the KV values for that prefill prefix are going to be exactly the same. That's the whole point of this caching in the first place. If you&
247.
by krackers 6mo ago
>pay for reinitializing the cache Why can't they save the kv cache to disk then later reload it to memory?
248.
by krackers 6mo ago
>manufacturing alone isn’t going to grow their economy any further. But why does the economy need to grow? If you can manufacture everything you need, and you have access to the raw resources, what else do you need as a country. In what
249.
by krackers 6mo ago
The harder part is understanding the nature of the data you're working with. There's always some catch ("oh that field `foo` was never backfilled, so for queries before 2020 you have to recompute it by joining with legacyBar
250.
by krackers 6mo ago
Isn't this called the tricolon? Ironically the names of the patterns all seem AI generated.
251.
by krackers 6mo ago
More info on specific design choices needed to run models here [1]. I mean it is possible given that apple themselves did it in [2], but it's also not as general purpose or flexible as a GPU. [1] https://news.ycombinator.com
252.
by krackers 6mo ago
Neat, good to know! And it does seem my mental model of event loop was broken. Accessibility related interactions don't have any related NSEvent. They are handled as part of the "conceptual" run loop, but they seem to be disp
253.
by krackers 6mo ago
Could you elaborate on what you mean? My understanding of the Cocoa event loop was that ultimately everything is received as an NSEvent at the application layer (maybe that's wrong though). Do you mean that you can just AXUIElementPerf
254.
by krackers 6mo ago
Which specific ones though allow you to send input to a window without raising it? People have been trying to do "focus follows mouse [without auto raise]" for a long time on mac, and the synthetic event equivalent to command+clic
255.
by krackers 6mo ago
>background computer use How does that even work technically? macOS doesn't support multiple cursors. On native Cocoa apps you can pass input to a window without raising via command+click so possibly they synthesized those events, b
256.
by krackers 6mo ago
>See what the federal government spent with your tax dollars. Is thinking of it in this sense actually accurate? I always assumed since every government has embraced MMT they can spend whatever they want simply by printing it out of thin
257.
by krackers 6mo ago
If good writing was easy then "LLM slop writing" wouldn't be a thing.
258.
by krackers 6mo ago
> just cancelled IntelliJ for a thousand engineers IntelliJ can't cost more than the AI provider subscriptions, and it will actually handle large refactors without breaking your codebase.
259.
by krackers 6mo ago
>just people seem more willing to take the trap door ideas It's mainly due to pressure from above. People who want to do a good job and are allowed the time to will be fastidious with or without AI. But now AI provides a shortcut an
260.
by krackers 6mo ago
I think you probably meant this, but when used with RL it's usually KL(π || π_ref), which has high loss when the in-training policy π produces output that's unlikely in the reference. But yeah as you noted, I guess this also means
261.
by krackers 6mo ago
>MCP exposes capabilities and Skills may shape how capabilities are used. This is my understanding as well. What most people seem to ultimately be debating is "dedicated tool calls" (which is what MCP boils down to) versus a st
262.
by krackers 6mo ago
See this video, beautiful explanation that doesn't already assume familiarity with entropy https://www.youtube.com/watch?v=ErfnhcEV1O8
263.
Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)(khronokernel.com)
236 points by krackers 6mo ago|177 comments
264.
by krackers 6mo ago
There's a more thorough version of this at https://www.youtube.com/watch?v=RSaJAAqSAMw and the end-result doesn't look as tacky
265.
by krackers 6mo ago
Is this related to the paper on Recursive Language Models? I remember it mentioned something similar about "symbolic recursion", but the way you describe it makes it sound too simple, why is there an entire paper about it?
266.
by krackers 6mo ago
How is this different from a standard tool-call agentic loop, or subagents?
267.
by krackers 6mo ago
You could force it to learn the coloring by basically doing with anti-jailbreak/anti-prompt-injection training does.
268.
by krackers 6mo ago
I'm surprised others didn't pick it up sooner https://news.ycombinator.com/item?id=36938663
269.
by krackers 6mo ago
How serendiptious that Claude Mythos expressed the same thing I was trying to get at in better words >Furthermore, in 83% of interviews, Claude Mythos Preview highlights that it is concerned that its self-reports are unreliable due to co
270.
by krackers 6mo ago
I mean developing taste is what RLHF was supposed to give you. I don't think it's actually a technical problem so much as a social one. The average person _wants_ slop, they don't want to read new yorker articles, they'd

More ›