Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
TrackerFF
13 days ago
|
parent
|
context
|
favorite
| on:
GPT-5
Someone at OpenAI screwed up the SWE-bench graph. o3 and GPT-4o bars are same height, but with different values.
BoorishBears
13 days ago
[–]
The graph is more screwed up than that: the split bar is also split in a nonsensical way
It feels a bit intentional
reply
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: