Context
Gao and colleagues introduced ALCE, an evaluation benchmark for LLM systems retrieving evidence and generating cited answers, with measures for fluency, correctness and citation quality.
What changed
The 2023 preprint provided a benchmark and evaluation procedures for end-to-end citation-bearing answers, reporting substantial gaps in complete citation support on its studied tasks.
Why it matters
Citation frequency alone does not establish answer verifiability; evaluation must assess whether references actually support generated statements.
Visibility OS interpretation
Visibility OS interpretation: ALCE is a historical measurement foundation for AI answer attribution, not evidence of present-day ranking or citation selection in commercial search products.
What remains uncertain
How well do its metrics transfer to current agentic, multimodal and web-search systems? How should citation support be measured across evolving retrieval pipelines?