Skip to content

SWE-bench history check: count PR mentions only since the PR was opened - #17

Merged
rzonyx merged 1 commit into
mainfrom
swebench-history-probe
Oct 28, 2026
Merged

rzonyx merged 1 commit into
mainfrom
swebench-history-probe

Conversation

@rzonyx

@rzonyx rzonyx commented Oct 23, 2026 •

Copy link
Copy Markdown
Contributor

The history check counted every commit message with #<number> as naming the pull request. Django cites Trac tickets the same way, so django__django-15315 failed on a correctly pruned history: its base keeps "Fixed #15315" from 2011, ticket 15315, not PR 15315 (2022).

  • The probe passes the task's created_at to git log --since: a commit from before the PR was opened can't refer to it.
  • Multi-SWE-bench tasks have no created_at and keep counting every message.

Tests: 3 new probe tests against a local git repository; unit tests pass, and django__django-15315 passed in #16's Sample tasks run with this commit.

The check counted every commit message with "#<number>" as naming the
pull request. Django's messages cite Trac tickets the same way, and ticket
15315 (2011) is not pull request 15315 (2022), so
django__django-15315 failed with a base history that is correctly pruned.
A commit from before the pull request was opened cannot refer to it, so
the probe now passes the task's created_at to git log --since.
Multi-SWE-bench tasks carry no created_at and keep counting every message.
@rzonyx
rzonyx merged commit 427b377 into main Oct 28, 2026
45 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant