Leijie Wang
@leijiew.bsky.social
A third-year PhD student at the University of Washington
Huge thanks to my wonderful collaborators Kathryn Yurechko, Pranati Dani, @cqz.bsky.social and @axz.bsky.social.
Full details here ➡️ arxiv.org/pdf/2409.03247
Full details here ➡️ arxiv.org/pdf/2409.03247
arxiv.org
March 25, 2025 at 1:09 AM
Huge thanks to my wonderful collaborators Kathryn Yurechko, Pranati Dani, @cqz.bsky.social and @axz.bsky.social.
Full details here ➡️ arxiv.org/pdf/2409.03247
Full details here ➡️ arxiv.org/pdf/2409.03247
All three strategies struggled with iterative refinement.
Interestingly, participants adopted hybrid approaches when iterating on their prompt filters – like providing examples as in-context examples or writing rule-like prompts.
Interestingly, participants adopted hybrid approaches when iterating on their prompt filters – like providing examples as in-context examples or writing rule-like prompts.
March 25, 2025 at 1:07 AM
All three strategies struggled with iterative refinement.
Interestingly, participants adopted hybrid approaches when iterating on their prompt filters – like providing examples as in-context examples or writing rule-like prompts.
Interestingly, participants adopted hybrid approaches when iterating on their prompt filters – like providing examples as in-context examples or writing rule-like prompts.
Despite 🤖LLM prompting’s better performance, participants preferred mixed strategies to create their filters.
For example, when their preferences were ill-defined but intuitive, 🔎labeling examples was considered the easiest way. (🧵4/N)
For example, when their preferences were ill-defined but intuitive, 🔎labeling examples was considered the easiest way. (🧵4/N)
March 25, 2025 at 1:07 AM
Despite 🤖LLM prompting’s better performance, participants preferred mixed strategies to create their filters.
For example, when their preferences were ill-defined but intuitive, 🔎labeling examples was considered the easiest way. (🧵4/N)
For example, when their preferences were ill-defined but intuitive, 🔎labeling examples was considered the easiest way. (🧵4/N)
To answer this question, our study had 37 non-programmers create personal content filters using these three strategies. (🧵3/N)
March 25, 2025 at 1:07 AM
To answer this question, our study had 37 non-programmers create personal content filters using these three strategies. (🧵3/N)
Existing content filter tools often expect lay people to work in a single, long setup session. Yet users engage with social media in short, everyday sessions.
How can we support social media users to more easily create and iterate on their filters? (🧵2/N)
How can we support social media users to more easily create and iterate on their filters? (🧵2/N)
March 25, 2025 at 1:07 AM
Existing content filter tools often expect lay people to work in a single, long setup session. Yet users engage with social media in short, everyday sessions.
How can we support social media users to more easily create and iterate on their filters? (🧵2/N)
How can we support social media users to more easily create and iterate on their filters? (🧵2/N)