FA-98539 / Beam search bookkeeping / Member archive

Beam search: beam stopping ignores the best normalized score a live path can still reach · case 04

beam stopping ignores the best normalized score a live path can still reach.

Member previewVariant 4 · 3 implementations · 8 checks per implementation

Case contract

Scores are nonpositive upper bounds on cumulative log scores; future token contributions cannot increase them. Active rows are [score_bound,current_generated_length], finished contains already-normalized final scores, width is positive, and max_length is at least each active length. The length factor is 1+alpha*(length-1), alpha>=0. Stop if there are no active paths, or at least width finished paths exist and the best active upper bound normalized at max_length is no better than the width-th best finished score.

Why this case matters

Beam decoders rely on consistent hypothesis state while expanding and ranking candidates.

One recorded failure

Sample boundary fixture

This sample comes from the broken implementation of a controlled reproducer.

Boundary fixture	Actual	Expected	Outcome
extension can improve rank	`true`	`false`	Failed

MEMBER ARCHIVE

The complete case is available to members.

This record includes three runnable implementations, regression fixtures, execution results, and source hashes.

Member access is invitation-based. Sign in with your invited account to inspect the sources.