Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
rtsil
on May 14, 2025
|
parent
|
context
|
favorite
| on:
AlphaEvolve: A Gemini-powered coding agent for des...
Maybe the actual solution to the interpretability/blackbox problem is to not ask the llm to execute a given task, but rather to write deterministic programs that can execute the task.
physix
on May 15, 2025
[–]
That is what I think is most interesting about it. You get repeatable efficiency gains rather than burning GPU time in data centres.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: