jdnier

5 ms·

31.
by jdnier 11mo ago
Yesterday there was a somewhat similar DuckDB post, "Frozen DuckLakes for Multi-User, Serverless Data Access". https://news.ycombinator.com/item?id=45702831
32.
Researchers find adding simple sentence to prompts makes AI models more creative(venturebeat.com)
6 points by jdnier 1y ago|0 comments
33.
Verbalized Sampling: How to Mitigate Mode Collapse and Unlock LLM Diversity(arxiv.org)
3 points by jdnier 1y ago|0 comments
34.
by jdnier 1y ago
Looking up the etymology of "sargeable", I found this StackOverflow answer: https://dba.stackexchange.com/a/217983 And Google explains "The term 'sargable' is a portmanteau of "Search ARGu
35.
Mysterious cosmic 'dots' are baffling astronomers. What are they?(nature.com)
16 points by jdnier 1y ago|1 comments
36.
by jdnier 1y ago
If you want to do this rigorously, I suggest you read Robert D. Cameron's excellent paper "REX: XML Shallow Parsing with Regular Expressions" (1998). https://www2.cs.sfu.ca/~cameron/REX.html
37.
by jdnier 1y ago
That must be for "Dissolution and crystallization of cobalt, copper and sodium chlorides". It's quite something to watch! https://www.nikonsmallworld.com/galleries/2025-small-world-i...
38.
by jdnier 1y ago
> should buy the books Yes I totally will, err..., oh my, ebook for $91.99, paperback for $127.99. What's going on with these prices? These aren't college textbooks. I'm glad to hear about the 3rd edition but the cost give
39.
by jdnier 1y ago
The paper they've submitted goes into a lot more detail. https://royalsocietypublishing.org/doi/10.1098/rspa.2025.029...
40.
by jdnier 1y ago
And it has a Python wrapper, https://github.com/argolab/dyna3/tree/master/dyna_python_mod... .
41.
by jdnier 1y ago
Wait, so this object is going to pass quite close to Jupiter, Mars, and Venus? And the Juno probe, which was planned to be decommissioned soon anyway, could potentially be maneuvered to follow and get close to the object? Kind of uncanny an
42.
New asteroids spotted within the Rubin Observatory's first hours(abc.net.au)
4 points by jdnier 1y ago|0 comments
43.
by jdnier 1y ago
Hi, just wondering what you're thinking about how your tool might be abused.
44.
We should protect the high seas from all extraction, forever(nature.com)
5 points by jdnier 1y ago|0 comments
45.
by jdnier 1y ago
An overview of other studies investigating cysteine and glycine sources and effects (from 2023; video or transcript): https://nutritionfacts.org/hnta-video/how-to-get-less-cystei...
46.
by jdnier 1y ago
LLMs trained on LLM generated material trained on ... until just gray goo is left.
47.
Materialized View Strategies Using PostgreSQL (2015)(hashrocket.com)
2 points by jdnier 1y ago|0 comments
48.
by jdnier 1y ago
So I wonder how much of Claude's perceived personality is due to the system prompt versus the underlying LLM and training. Could you layer a "Claude mode"—like a vim/emacs mode—on ChatGPT or some other LLM by using a sim
49.
by jdnier 1y ago
Make sure the check out the LLM chat log he shared for the current article! The tone is cynical HN commenter indeed. https://github.com/minimaxir/llm-use/blob/main/criticism_hn....
50.
by jdnier 1y ago
> Alright! We managed to plot two sine waves in the most convoluted method ever! Those sine waves start to resemble a multi-voice music score. You might get some interesting pitch variability depending on other processes that are running
51.
by jdnier 1y ago
The author is Co-founder of Databricks, creator of K Prize, so an early adopter.
52.
How to improve DuckDB performance on Parquet(mooncake.dev)
2 points by jdnier 2y ago|0 comments
53.
by jdnier 2y ago
Interesting article relating Adam Unikowsky asking for Claude to decide a Supreme Court case: https://blog.plover.com/tech/gpt/presidential-emoji.html "Claude is fully capable of acting as a Supreme Court Jus
54.
by jdnier 2y ago
See "Rye and Uv: August Is Harvest Season for Python Packaging"[1] [1] https://lucumr.pocoo.org/2024/8/21/harvest-season/
55.
How Soon Might the Atlantic Ocean Break? Two Sibling Scientists Found an Answer(wired.com)
61 points by jdnier 2y ago|35 comments
56.
by jdnier 2y ago
Are you able to describe what makes them better? (Honest question, I'm not familiar with either or with Amphi.) It seems Definite's use case is focused on connecting to lots of data sources. For much smaller scale, how does Amphi
57.
by jdnier 2y ago
Wow, django-ninja is just what I was looking for. Thanks for pointing it out. Here’s their CRUD example: https://django-ninja.dev/tutorial/other/crud/
58.
Canadian science gets biggest boost to PhD and postdoc pay in 20 years(nature.com)
2 points by jdnier 2y ago|0 comments
59.
by jdnier 3y ago
Thanks for pointing this out. I had never really thought much about it. I 100% agree with you! The docs are totally why Python is so comfortable for me.
60.
Welcome to the World of Statecharts(statecharts.dev)
3 points by jdnier 3y ago|0 comments

More ›