The Attribution-Compression Frontier in Retrieval-Augmented Generation
Quick summary
arXiv:2609.14245v1 Announce Type: cross Abstract: Context compression reduces generator input in retrieval-augmented generation, but answer quality alone does not characterize citation attribution. We measure citation attribution across compression methods and budgets, comparing reranking, extractive selection, abstractive summarization, token pruning, and an extract-cluster-rewrite construction on ASQA and QASPER under a fixed generator and primary entailment evaluator. On ASQA at a nominal 0.25 budget (achieved compression 0.08), a RECOMP-style compressor's citations score 0.86 precision aga
Key takeaways
- arXiv:2609.14245v1 Announce Type: cross Abstract: Context compression reduces generator input in retrieval-augmented generation, but answer quality alone does not characterize citation attribution.
- We measure citation attribution across compression methods and budgets, comparing reranking, extractive selection, abstractive summarization, token pruning, and an extract-cluster-rewrite construction on ASQA and QASPER under a fixed generator and primary entailment evaluator.
- On ASQA at a nominal 0.25 budget (achieved compression 0.08), a RECOMP-style compressor's citations score 0.86 precision aga
Why it matters
“The Attribution-Compression Frontier in Retrieval-Augmented Generation” illustrates how changes in the AI ecosystem can affect products, workflows and user expectations together. Its lasting significance depends on measurable adoption, cost and safety outcomes.
Member comments