Skip to content

package: mean_score over full benchmark (not attempted) #48

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
john-b-yang merged 1 commit into from
Jun 22, 2026
Merged

package: mean_score over full benchmark (not attempted) #48

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions src/programbench/submission.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -183,10 +183,10 @@ def aggregate(scores: dict[str, float], n_total: int) -> Headline:
if not values:
raise ValueError("No scored instances found")
n = len(values)
# mean is over attempted instances; resolved/near are over the full benchmark
# (an unattempted task counts as unresolved).
# mean, resolved, and near are all over the full benchmark — an unattempted task counts
# as 0, matching how the leaderboard scores partial submissions.
return Headline(
mean_score=round(sum(values) / n, 4),
mean_score=round(sum(values) / n_total, 4),
resolved_pct=round(100 * sum(s >= RESOLVED_THRESHOLD for s in values) / n_total, 1),
near_resolved_pct=round(100 * sum(s >= NEAR_RESOLVED_THRESHOLD for s in values) / n_total, 1),
n_instances_attempted=n,
Expand Down
Loading