modified L10N prompt that provides significantly better outcomes #6893
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
escattone
changed the title
akatsoulas
approved these changes
Sep 5, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
mozilla/sumo#2513
mozilla/sumo#2514
mozilla/sumo#2516
Notes
Locally, I made 20 translation runs using the old prompt, and 20 runs using the new prompt, using the same content as this KB article revision.
Using the old prompt, I discovered that the LLM sometimes ignored the rules forbidding the translation of
menu/button/preftags and Wiki article link titles (e.g.,[[Some Article Title]]). In fact, in my series of runs, the rule forbidding the translation ofmenu/button/preftags and their content was commonly ignored (~30-40% of the time)After carefully reading the LLM's explanations from all of the runs, it seems that one of the sources of its "mistakes" in following the instructions -- maybe even the primary source -- is that it sometimes thinks its role as an expert translator supersedes the explicit instructions that we've provided in the prompt.
Using the prompt changes/additions included in this PR -- the most important being the addition that informs the LLM that the rules/instructions within the prompt take precedence over its role as an expert translator -- all 20 runs completed without any mistakes in following the rules/instructions.