zhimin-z
commited on
Commit
·
be4d5d9
1
Parent(s):
591e5b2
refine
Browse files
README.md
CHANGED
|
@@ -19,7 +19,7 @@ No benchmarks. No sandboxes. Just real issues that got resolved.
|
|
| 19 |
|
| 20 |
## Why This Exists
|
| 21 |
|
| 22 |
-
Most AI
|
| 23 |
|
| 24 |
If an assistant can consistently resolve issues across different projects, that tells you something no benchmark can.
|
| 25 |
|
|
|
|
| 19 |
|
| 20 |
## Why This Exists
|
| 21 |
|
| 22 |
+
Most AI assistant benchmarks use synthetic tasks and simulated environments. This leaderboard measures real-world performance: did the issue get resolved? How many were completed? Is the assistant improving?
|
| 23 |
|
| 24 |
If an assistant can consistently resolve issues across different projects, that tells you something no benchmark can.
|
| 25 |
|