Hacker News: gliptic

New comment by gliptic in "OpenAI and Hugging Face address security incident during model evaluation"

gliptic — Fri, 24 Jul 2026 21:39:19 +0000

The very wikipedia page you're linking says:

> The term abliteration has been coined for the process of using ablation to uncensor large language models by modifying internal functions to completely eliminate refusal behaviors while preserving the remaining functions of the model. The word is a portmanteau that combines the words ablation and obliteration.

So are you dropping this now?

New comment by gliptic in "OpenAI and Hugging Face address security incident during model evaluation"

gliptic — Fri, 24 Jul 2026 16:00:14 +0000

https://en.wiktionary.org/wiki/abliterate

New comment by gliptic in "OpenAI and Hugging Face address security incident during model evaluation"

gliptic — Fri, 24 Jul 2026 15:58:00 +0000

Except we're talking about a _different_ word, a word that was _coined_ as a portmanteau of the word you're referring to and another word.

New comment by gliptic in "OpenAI and Hugging Face address security incident during model evaluation"

gliptic — Wed, 22 Jul 2026 09:18:34 +0000

The name for it _is_ abliterate. It's a portmanteau of ablate and obliterate.

New comment by gliptic in "Zig ELF Linker Improvements Devlog"

gliptic — Sat, 30 May 2026 20:40:23 +0000

> Zig-native immediate-mode

dvui?

New comment by gliptic in "Rust--: Rust without the borrow checker"

gliptic — Thu, 01 Jan 2026 12:51:19 +0000

The borrow checker doesn't decide when things are dropped. It only checks reference uses and doesn't generate any code. This will work exactly the same as long as your program doesn't violate any borrowing rules.

New comment by gliptic in "Roomba maker goes bankrupt, Chinese owner emerges"

gliptic — Mon, 15 Dec 2025 09:42:30 +0000

Camping is like.. regulatory capture? Stretching the analogy thin here.

New comment by gliptic in "A small number of samples can poison LLMs of any size"

gliptic — Thu, 09 Oct 2025 19:17:15 +0000

But that fine-tuning is done only on those 100-200 good samples. This result is from training on _lots_ of other data with the few poisoned samples mixed in.

New comment by gliptic in "Cat Aquariums"

gliptic — Sun, 14 Sep 2025 15:45:23 +0000

This doesn't really make sense to me. Most cats I've known react to such reflections without ever having seen a laser pointer in their life, for the same reason they react to laser pointers.

New comment by gliptic in "Two Slice, a font that's only 2px tall"

gliptic — Sun, 14 Sep 2025 10:00:18 +0000

Glyph advance or line spacing is not part of the bitmaps.

New comment by gliptic in "Thai Air Force seals deal for Swedish Gripen jets"

gliptic — Sat, 16 Aug 2025 06:26:06 +0000

> I think Sweden was deploying the Drakken (Dragon) and later the Vigen (Lightning).

The names are much less flashy, Draken (The kite, due to the shape) and Viggen (The tufted duck) :P.

New comment by gliptic in "There is no memory safety without thread safety"

gliptic — Thu, 24 Jul 2025 18:25:46 +0000

`f(++i, ++i)` is/was indeed UB, but the example in munificent's comment was `foo(print(1), print(2))` which as far as I know is not even if both `print` calls read/write the same memory.

New comment by gliptic in "There is no memory safety without thread safety"

gliptic — Thu, 24 Jul 2025 17:54:49 +0000

Yes, but that's just a subset of expressions where unspecified sequencing applied. For instance, the example with two `print()` as parameters would have a sequence point (in pre-C++11 terminology) separating any reads/writes inside the `print` due to the function calls. It would never be UB even though the order in which the prints are called is still unspecified.

New comment by gliptic in "There is no memory safety without thread safety"

gliptic — Thu, 24 Jul 2025 17:34:23 +0000

The author is using the term in the way that everyone else understands it. They are not aware of your unusual definition.

New comment by gliptic in "There is no memory safety without thread safety"

gliptic — Thu, 24 Jul 2025 17:29:14 +0000

The evaluation order is _unspecified_, not undefined behaviour.

New comment by gliptic in "BB(6) Is Hard (Antihydra) (2024)"

gliptic — Thu, 17 Jul 2025 13:43:45 +0000

Recent developments on BB(6) previously posted here: https://scottaaronson.blog/?p=8972

New comment by gliptic in "Zig, the Ideal C Replacement Or?"

gliptic — Sun, 11 May 2025 17:02:30 +0000

It's really not. Looks like something like JS with almost identical syntax.

Also, the compiler (and what else?) isn't even implemented [1].

[1] https://github.com/profullstack/smashlang/blob/master/src/co...

New comment by gliptic in "Zig: A new direction for low-level programming?"

gliptic — Sun, 11 May 2025 12:41:35 +0000

> Dynamic Typing: Flexible type system with runtime type checking

Uhm, what does this have to do with C/Zig?

New comment by gliptic in "Llama 4 Smells Bad"

gliptic — Thu, 24 Apr 2025 08:47:23 +0000

Ah, FastML is an extremely overloaded name.

New comment by gliptic in "Llama 4 Smells Bad"

gliptic — Thu, 24 Apr 2025 07:44:13 +0000

It's strange that someone from FastML can be confused about this, unless it's supposed to be a bad joke.