Weekly Update

• Tone Check edit check
  • Discussing how to incorporate user feedback
  • Designing the structured task
  • Improving the Model card

Weekly Update

• Moderator recommendation: No updates
• Tone Check edit check: Supporting ML and Product on designing structured task around tone check.
• Edit diffs: @fkaelin to update.

Weekly update for edit diffs:

• Notebook for exploring query patterns.
• Switched to using edit types instead of tokenizing the added/removed words using the mwtokenizer. The structured output of edit-types is much preferable to imposing structure on the aggregated raw tokens. Running edit-types at scale is computationally challenging due the memory hungriness of the parsing library (mwparserfromhell). The approach to have a daily dag that appends to an iceberg table seems to work, likely because the suspected memory leaks don't have "enough time" to explode the job, as each batch only computes one day of data. This will be helpful for T351225.

Weekly Update

• Tone Check edit check: Supporting team with the slide preparation for the staff meeting presentation.
• Moderator recommendation: Working on the experiment design.
  • Split editors by user group and finding recommendation for specific group.
  • This drastically decrease the number of users (and data sparsity)

Weekly report

• Edit diffs:
  • a prototype for querying editor history is deployed on cloud vps, with a design plan document.
• Tone check:
  • Represented research at Tone Check product presentation at All Staff meeting
  • Met with (part of) the Tone Check team to discuss how to incorporate user feedback from the future Tone Check structured task.
    • We need to validate newcomers expertise before consider their feedback to retrain the model.

Weekly report

• Moderator recommendation: Work in selecting the task to recommend and categorize them according to their predictability.

Weekly Report

• Edit diffs: collected feedback about the prototype from product management.