# EviRerank: Adaptive Evidence Construction for Long-Document LLM Reranking

arXiv: https://arxiv.org/abs/2411.06254 · HTML: https://mlsanity.com/paper/2411.06254/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 9 Nov 2024.

## Built on by (most cited first, 3 of 3)

- [Retrieval-Feedback-Driven Distillation and Preference Alignment for Efficient LLM-based Query Expansion](https://mlsanity.com/paper/2603.13776/) (2026)
- [W-RAG: Source-Aware Retrieval for Enterprise Document Generation from Heterogeneous Knowledge Bases](https://mlsanity.com/paper/2608.22081/) (2026)
- [BoundaryMORPH: Budgeted Reranking via Active Set Selection for Diffuse Retrieval](https://mlsanity.com/paper/2609.27213/) (2026)

## Abstract (arXiv)

Decoder-only LLM rerankers struggle with long documents: inference is costly and relevance signals can be diluted by irrelevant context. Motivated by an attention analysis indicating a consistent degradation trend when non-relevant text is appended, we propose EviRerank, an evidence-based long-document reranking framework for decoder-only LLMs. EviRerank (i) scores document blocks with a lightweight selector (BM25, bi-encoder, or cross-encoder), (ii) constructs a compact reranking context under a hard token cap by dynamically budgeting evidence blocks with Adaptive Evidence Budgeting (AEB) and adding a global summary cue via Summary Augmentation (SA), and (iii) reranks with a decoder-only LLM. Across TREC DL'19, DL'22, DL'23, and MLDR-zh, EviRerank consistently outperforms full-document LLM reranking and strong block-selection baselines while substantially reducing the required input length. On TREC DL'19, EviRerank reaches up to 0.744 nDCG@10 while substantially reducing reranking latency. Additional RankZephyr-7B experiments confirm transfer to listwise reranking under matched context budgets.