HackerTrans
トップ新着トレンドコメント過去質問紹介求人

ajoy

no profile record

投稿

Prompt Repetition Improves Non-Reasoning LLMs

arxiv.org
1 ポイント·投稿者 ajoy·5 か月前·1 コメント

The Looming Taiwan Chip Disaster That Silicon Valley Has Long Ignored

nytimes.com
4 ポイント·投稿者 ajoy·5 か月前·0 コメント

How to Remember What You Read

fs.blog
330 ポイント·投稿者 ajoy·5 年前·91 コメント

コメント

ajoy
·4 か月前·議論
Reminded me of a similar story at openSSH, wonderfully documented in a "Veritasium" episode, which was just fascinating to watch/listen.

https://www.youtube.com/watch?v=aoag03mSuXQ
ajoy
·5 か月前·議論
"When not using reasoning, repeating the input prompt improves performance for popular models (Gemini, GPT, Claude, and Deepseek) without increasing the number of generated tokens or latency."
ajoy
·5 年前·議論
I think we have to consider the possibility (most likely), that the advanced versions of these lifeforms are much further along the evolutionary tree than we are. They might have evolved like us, but continued evolving and now look nothing like us.
ajoy
·6 年前·議論
Most people don't know this, but Nokia had hit upon a lot of these ideas before the iPhone, but didn't act on it : https://venturebeat.com/2010/09/27/nokia-had-iphone-like-pro...

They were an innovative company, till internal culture or complacency killed it (conjecture).