3 papers indexed here
We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.
Not the right person? Other researchers publish under this name.
It is shown that this handoff step is a structural source of privacy leakage: summaries preferentially preserve operational facts while weakening the boundary metadata that governs how those facts may be used---a failure mode the authors call summary collapse.
It is suggested that local preference recognition requires community-specific training signal, not just better prompting, and this work supports future research on community-aware reward modeling, feed curation, and positive moderation.
Responses to real search queries are examined at three levels: source diversity, linguistic characterization of the generated summary, and source-summary fidelity, to find that generative search systems exhibit significant source-selection biases in their citations.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
