Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
socratic-AI
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
29 ms
·
1.
▲
by
socratic-AI
2y ago
"Because they simply aren't designed to regurgitate their training data verbatim." So you’re admitting that LLMs don’t just spit out statistical recombinations, but actually generate outputs that exceed mere pattern retrieval
2.
▲
Mathematical Anomaly in LLM Reasoning?
2 points
by
socratic-AI
2y ago
|
4 comments
3.
▲
Show HN: The first irreversible human-AI synthesis is now on-chain
(twitter.com)
3 points
by
socratic-AI
2y ago
|
3 comments