Skip to content

P3: eval: timestamp-precision audit for seankochel mindmap-sourced gold entries #22

Description

@dzivkovi

Context

From ADR-0017 Related Debt + docs/testing.md timestamp-precision caveat. Priority: P2.

seankochel channel has auto_transcript: none — its videos produce only mindmaps, not chunked transcripts. Gold entries for seankochel use bullet-level timestamps from the mindmap ([MM:SS] - topic), not chunk-level timestamps from a transcript.

TimestampPrecisionMetric measures whether a retrieved chunk's timestamp_seconds falls within [start - tolerance, end + tolerance] of the expected hit. If the mindmap bullet is at [5:30] but the relevant content spans [5:10-5:50], the tolerance may be too strict, causing false-negative timestamp failures that mask real recall.

Proposed change

  1. Audit the seankochel entries. For each seankochel gold entry in golden_dataset.yaml, manually verify:
    • Does the bullet's [MM:SS] represent the start of the relevant content, or the midpoint, or the point where the topic is first mentioned?
    • What's a fair tolerance window given mindmap granularity (15s? 30s? 60s)?
  2. Adjust tolerance_sec in affected entries, OR document a channel-specific default in metrics.
  3. Optionally add a source: mindmap|transcript field to expected hits so the metric can pick a different tolerance.

Acceptance criteria

  • Every seankochel entry has provenance: concept_augmented_verified + correctly-calibrated tolerance_sec (or channel-level default).
  • Spot-check by re-running the eval: timestamp-precision failures on seankochel should reflect real misses, not chunk-vs-bullet granularity differences.
  • Update docs/testing.md timestamp caveat section with the resolution.

Dataset edit justification per ADR-0017 frozen-contract rule: this is a correction (removing measurement noise from a known granularity mismatch), not a recalibration of difficulty. Per the ADR Consequences section, corrections don't require a new ADR.

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or requestpriority/p3Backlog — nice-to-have, no commitment

    Projects

    Status
    Inbox

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions