Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
glth
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
1.
▲
by
glth
7mo ago
Author here: I have just fixed the typo. Thank you. And yes, these are screenshots from Anthropic’s documentation.
2.
▲
Why XML tags are so fundamental to Claude
(glthr.com)
228 points
by
glth
7mo ago
|
153 comments
3.
▲
by
glth
7mo ago
On a related note, I have been experimenting with a small prototype for cross-agent, device-local active memory called brAIn ( https://github.com/glthr/brAIn ). It delivers a personalized agent experience with everything
4.
▲
BrAIn: Persistent, Human-Inspired Memory for LLM Agents
(github.com)
2 points
by
glth
7mo ago
|
0 comments
5.
▲
by
glth
9mo ago
https://glthr.com/
6.
▲
The Riven Diffs – Seeing Riven (1997) Differently
(glthr.com)
81 points
by
glth
9mo ago
|
11 comments
7.
▲
Formulaic Delimiters in the Iliad and the Odyssey
(glthr.com)
30 points
by
glth
9mo ago
|
13 comments
8.
▲
Apocolocyntosis
(en.wikipedia.org)
1 points
by
glth
10mo ago
|
0 comments
9.
▲
Roman Lead Pipe Inscription
(en.wikipedia.org)
1 points
by
glth
1y ago
|
1 comments
10.
▲
Model Collapse and the Need for Human-Generated Training Data
(glthr.com)
16 points
by
glth
1y ago
|
3 comments
11.
▲
Software is evolving backwards [video]
(youtube.com)
4 points
by
glth
1y ago
|
1 comments
12.
▲
by
glth
2y ago
Author here: good catch, thank you! I have fixed the two errors.
13.
▲
by
glth
2y ago
Author here: In a few weeks, I will publish the graph generator's source code and a DOT file encoding the Myst graph in this repository: https://github.com/glthr/DeMystify . I would love to see how others render th
14.
▲
by
glth
2y ago
Interesting: I would love to read this article when it is published.
15.
▲
by
glth
2y ago
I think that is a fantastic idea! I would be happy to see your blog take shape. Blogging is something I genuinely enjoy, as it allows me to put my thoughts into tangible form.
16.
▲
The Myst Graph: A New Perspective on Myst
(glthr.com)
3 points
by
glth
2y ago
|
0 comments
17.
▲
by
glth
2y ago
Great video! Very instructive.
18.
▲
Discovering errors in Donald Knuth's TAOCP
(glthr.com)
207 points
by
glth
2y ago
|
65 comments
19.
▲
by
glth
2y ago
I am currently exploring methods to detect and categorize undocumented special tokens (or: "delimiters") used by LLMs. As part of this effort, I have developed a preliminary prototype (named deLLMiter) that I am refining. This met
20.
▲
LLM Delimiters and Higher-Order Expressions
(glthr.com)
4 points
by
glth
2y ago
|
0 comments
21.
▲
Super Bright Anti-Submarine Searchlight [video]
(youtube.com)
1 points
by
glth
2y ago
|
0 comments
22.
▲
First-Order and Second-Order Expressions
(glthr.com)
2 points
by
glth
2y ago
|
0 comments
23.
▲
by
glth
2y ago
alias: G. interests: computers, programming languages, programmatic puzzles, bacteriophages language(s): en, fr link to something you think is cool: https://google-research.github.io/self-organising-systems/20
24.
▲
The Metamorphosis of Prime Intellect
(en.wikipedia.org)
2 points
by
glth
2y ago
|
0 comments
25.
▲
Graffito of Esmet-Akhom (last known hieroglyph inscription)
(en.wikipedia.org)
2 points
by
glth
2y ago
|
0 comments
26.
▲
Solving "Number Cross 4" (Jane Street Puzzle, May 2024)
(glthr.com)
1 points
by
glth
2y ago
|
0 comments
27.
▲
Show HN: K4nundrum – An open-source tool to investigate Kryptos K4
(github.com)
1 points
by
glth
3y ago
|
0 comments
28.
▲
Parietal Eye
(en.wikipedia.org)
2 points
by
glth
3y ago
|
0 comments
29.
▲
A fresh impression of Freud's 'mystic writing-pad'
(news.yale.edu)
2 points
by
glth
3y ago
|
0 comments
30.
▲
A Path Towards Autonomous Machine Intelligence
(openreview.net)
1 points
by
glth
3y ago
|
1 comments
More ›