Hacker Newsnew | past | comments | ask | show | jobs | submit | glth's commentslogin

Author here: I have just fixed the typo. Thank you.

And yes, these are screenshots from Anthropic’s documentation.


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 stored locally in a single file (agent.brain), and supports reusing semantic memory across projects. In practice, this means brAIn can identify and apply behavioral patterns you have used in other contexts whenever they are relevant. (I realize the repository should include a concrete example of this, and I will update it today to add one).



Author here: good catch, thank you! I have fixed the two errors.


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 the graph and how the generator can be improved. A follow-up article will also dive deeper into the technical details. Stay tuned!


I'm looking forward to it! Will try to share if it turns out well with what I have in mind.


Interesting: I would love to read this article when it is published.


...and here you go:

https://manuel.kiessling.net/2025/03/31/how-seasoned-develop...

Feedback more than welcome!


Same!



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.


Great video! Very instructive.


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 method is founded upon the hypothesis that there exists a correlation between first-order expressions and more complex, hierarchical forms of expression (https://glthr.com/llm-delimiters-and-higher-order-expression...).

With this project, my goal is to improve the security of LLM models.


  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/2022/diff-fsm/
  contact info: moeie.gyj/btfgxnc (Vigenère cipher, the key is this website's name)


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: