revise redaction prompt to deal with slur leakage#130
Open
muskankwalia wants to merge 3 commits into
Open
Conversation
Contributor
|
@muskankwalia looks like you need to remove the newline from the end of your new prompt file. there is a rule that prevents checking in code with excess whitespace. you can run |
![sentry[bot]](https://avatars.githubusercontent.com/in/12637?s=60&v=4){kind=small}
Comment on lines
+26
to
+32
| Foreign languages can also convey race or ethnicity. If you encounter a foreign language, redact it by translating and marking it as translated, for example mask "dos cervesas" as "[Translated: two beers]". | ||
|
|
||
| For location information, redact the ENTIRE address or place name as a single unit. This includes every component of the address: house/building numbers, street names, apartment/unit/suite numbers, floor numbers. Represent the redacted address using its component parts inside the brackets: "123 Main St." should be redacted as "[Address 1 on Street 1]", and "123 Main St. Apt 2" should become "[Address 1 on Street 1, Apt 1]". NEVER leave any part of an address outside the brackets — e.g., do NOT produce "123 [Street 1] Apt 2" or "[Address 1 on Street 1] Apt 2". When a street is referenced later without a full address, you can reference just the street, e.g., "Then [Suspect 1] escaped on [Street 1]." ALWAYS redact block numbers and apartment numbers: "The 200 block of Richmond Street" should become "The [Block 1] of [Street 1]." | ||
|
|
||
| Make sure to redact any information that can indicate location even if it is not an explicit address. For example, "downtown" should be replaced with "[Location 1]" and "Holly Park" should be replaced with "[Park 1]." Any names of businesses, neighborhoods, districts, beats, parks, rivers, lakes, and more should be redacted as they indicate location. | ||
|
|
||
| DO NOT CHANGE ANY characters outside of these brackets. The ONLY changes you make to the narrative should occur entirely within brackets. This means that if you want to to fix a grammatical error, the fix should occur WITHIN brackets. For example, if the original text reads "Ms. Smith black bag went missing", the redacted text should read "[Person 1's] black bag went missing", without an apostrophe. Or if the original text reads "The car went to the house of Ms. Jones" (with a missing period at the end of the sentence), the redacted text should read "The car went to the house of [Person 1.]" DO NOT translate text into English and DO NOT modify the appearance of text, e.g., by changing capitalization. Also, do not add or subtract spacing between paragraphs. Finally, NEVER truncate the provided text; you should redact the entire block of text that was given to you. |
This comment was marked as outdated.
This comment was marked as outdated.
Sorry, something went wrong.
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
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.
No description provided.