SilenN

4 ms·

31.
I spent 2 weeks playing god. My learnings from 597 genetic algorithm lineages(blog.silennai.com)
5 points by SilenN 8mo ago|2 comments
32.
The pragmatic tradeoff of tied embeddings(blog.silennai.com)
1 points by SilenN 8mo ago|1 comments
33.
by SilenN 8mo ago
Simply, it's when your output embedding matrix = input. You save vocab_dim*model_dim params (ex. 617m for GPT-3). But the residual stream means that the weight matrices are roughly connected via a matmul, which means they struggle to e
34.
by SilenN 8mo ago
https://news.ycombinator.com/item?id=46685327
35.
by SilenN 8mo ago
I disregard any comments like this one as baseless hate unless given an example or something constructive.
36.
by SilenN 8mo ago
See what you don't understand is that you need to coordinate the deacon to take the witness out back and talk to the mayor. It's actually quite trivial.
37.
by SilenN 8mo ago
Yes! I went to a meetup for it as well - https://x.com/silennai/status/1907540814890521023?s=20 - https://x.com/silennai/status/1918166890322784407?s=20
38.
by SilenN 8mo ago
This is probably true and not emphasized enough. Many are making their value judgements with token limits (which is fair, it is expensive).
39.
by SilenN 8mo ago
Good point. I'll consciously be trying to write more "out of distribution" now.
40.
by SilenN 8mo ago
I think you read my comment, but you didn't really READ it.
41.
by SilenN 8mo ago
See https://news.ycombinator.com/item?id=46685489 https://news.ycombinator.com/item?id=46687347
42.
by SilenN 8mo ago
That section doesn't include a lick of AI writing. One tell (maybe?) is that I switch from past tense to present tense mid sentence. Either a) I sound like an LLM when I'm writing articles (possible) or b) turing test AGI somethin
43.
by SilenN 8mo ago
Appreciate it! And agree. I added a longer comment above saying basically this.
44.
by SilenN 8mo ago
Outside of the story at the start I intentionally tried to make it information dense. Would appreciate feedback!
45.
by SilenN 8mo ago
Would appreciate feedback, see my comment above!
46.
by SilenN 8mo ago
What sections where you most off put by? I used Claude to expand on my ideas for a few of the purely informational things, and for formatting, but this article is largely written by hand. For example "Interface tests are the ability to
47.
by SilenN 8mo ago
Try Cursor Composer! It's the most natural transition. Exactly what you're currently doing, but it inserts the code snippets for you from within your IDE.
48.
by SilenN 8mo ago
How can I improve it in the future? The intention was to start as a story, and become more of a structured guide as you read on.
49.
by SilenN 8mo ago
The mayor doesn't communicate to the deacon or refiner well enough. Sometimes my polecat acts like a witness so I have to restart it...
50.
by SilenN 8mo ago
Nothing competes with Opus 4.5 in Claude Code. Codex & Gemini are arguable.
51.
by SilenN 8mo ago
https://github.com/SilenNaihin
52.
by SilenN 8mo ago
Yes, the technical preview of Github Copilot. I rounded up.
53.
by SilenN 8mo ago
Agree! Don't let it get too bad. See my other comment on debugging and refactoring being a core part of your workflow.
54.
by SilenN 8mo ago
^this
55.
by SilenN 8mo ago
Agree. And with the comments in the thread. I'll caveat my statement, with AI ready repos . Meaning those with good documentation, good comments (ex. avoiding Chestertons fence), comprehensive interface tests, Sentry, CI/CD, etc.
56.
by SilenN 8mo ago
Any suggestions on what to add to answer the question better? I tried to cover this in "Why I switched", "When to use Cursor", and "My current setup" sections.
57.
by SilenN 8mo ago
This. Also I'm quoting the email they sent me.
58.
I was a top 0.01% Cursor user, then switched to Claude Code 2.0(blog.silennai.com)
149 points by SilenN 8mo ago|230 comments