Skip to content

Update inference section with figure reference #76

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
jaseweston merged 1 commit into from
Mar 30, 2026
+1 −1
Merged

Update inference section with figure reference #76

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
2 changes: 1 addition & 1 deletion projects/paragator/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -107,7 +107,7 @@ Intuitively, the model is rewarded when at least one of its $m$ attempts solves

### Inference

During training, we optimize a single round of aggregation over one candidate pool. At inference time, however, we naturally generalize this to multiple iterations of aggregation, enabling sequential scaling in addition to the learned parallel sampling.
During training, we optimize a single round of aggregation over one candidate pool. At inference time, however, we naturally generalize this to multiple iterations of aggregation, enabling sequential scaling in addition to the learned parallel sampling (see inference figure above).

Concretely, given problem $x$, we first sample an initial pool of $m$ candidates $y_{1:m}^{(0)} \sim \mathcal{M}_\theta(y \mid x)$ and sample $m$ aggregated solutions

Expand Down
Loading