Skip to content

research: Add unseen-entity generalization experiment - #5

Merged
Samarthuday merged 1 commit into
mainfrom
research/unseen-entity-generalization
Aug 29, 2026
Merged

research: Add unseen-entity generalization experiment#5
Samarthuday merged 1 commit into
mainfrom
research/unseen-entity-generalization

Conversation

@Samarthuday

Copy link
Copy Markdown
Owner

Network features (out-degree, counterparty HHI, lifetime counts) drive most of the ablation lift (PR-AUC 0.09 -> 0.94), which raises an obvious question: does that hold for accounts the model never saw during training, or is it propped up by account history that only exists for already-observed entities?

scripts/unseen_entity_evaluation.py reuses the existing artifact (no retraining) and partitions the held-out test set by whether both sender and receiver appeared in the training window:

  • Standard out-of-time (full test set): PR-AUC 0.9859, lift 198.6x
  • Both parties seen during training: PR-AUC 0.9918, lift 198.6x
  • At least one party unseen: PR-AUC 0.7912, lift 198.4x

PR-AUC and precision drop sharply for unseen entities, but ROC-AUC (0.9998), recall @ 0.5% (~99.2%), and lift over base rate (~198x) are essentially unchanged across all three partitions. The PR-AUC drop tracks an 8x lower positive prevalence in the unseen-entity subset, not a loss of ranking ability -- PR-AUC's baseline is the prevalence itself, so a rarer-positive subset produces a lower score even under identical discrimination. Full writeup: docs/research/unseen_entity_generalization.md

Also notes what this does NOT test: a connected-component/network holdout (colluding clusters never seen together) is a stricter and still-open question, left as further work.

Network features (out-degree, counterparty HHI, lifetime counts) drive
most of the ablation lift (PR-AUC 0.09 -> 0.94), which raises an obvious
question: does that hold for accounts the model never saw during
training, or is it propped up by account history that only exists for
already-observed entities?

scripts/unseen_entity_evaluation.py reuses the existing artifact (no
retraining) and partitions the held-out test set by whether both
sender and receiver appeared in the training window:

- Standard out-of-time (full test set):  PR-AUC 0.9859, lift 198.6x
- Both parties seen during training:     PR-AUC 0.9918, lift 198.6x
- At least one party unseen:             PR-AUC 0.7912, lift 198.4x

PR-AUC and precision drop sharply for unseen entities, but ROC-AUC
(0.9998), recall @ 0.5% (~99.2%), and lift over base rate (~198x) are
essentially unchanged across all three partitions. The PR-AUC drop
tracks an 8x lower positive prevalence in the unseen-entity subset, not
a loss of ranking ability -- PR-AUC's baseline is the prevalence itself,
so a rarer-positive subset produces a lower score even under identical
discrimination. Full writeup: docs/research/unseen_entity_generalization.md

Also notes what this does NOT test: a connected-component/network holdout
(colluding clusters never seen together) is a stricter and still-open
question, left as further work.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Copilot AI lite review requested due to automatic review settings August 29, 2026 15:44

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

@Samarthuday
Samarthuday requested a lite review from Copilot August 29, 2026 15:44

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

@Samarthuday
Samarthuday merged commit c83455f into main Aug 29, 2026
2 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.

2 participants