For the best experience on desktop, install the
Chrome extension
to track your reading on news.ycombinator.com
×
Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
|
history
|
from
register
Simular Agent S hits 72.6% success on 369 real computer tasks (human: 72.36%)
(
os-world.github.io
)
2 points
by
taro666
4 months ago
|
past
OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computers
(
os-world.github.io
)
77 points
by
kristianpaul
on April 28, 2024
|
past
|
39 comments
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
×
HN For You
Display Mode
Highlight
Top
Only
Debug mode
Sign Out
API Key:
Connect
Create an account
to get your API key.