---
title: "Glance: The Generator Is Its Own Reranker"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2610.40454v1"
permalink: "https://feed.astralor.com/items/itm_022256a0b411f27fc6716a1d"
published: "2026-10-29T17:01:22.000Z"
score: 55
channel: "ai"
tags: ["RAG","Glance","The Generator","Its Own Reranker","Retrieval-augmented","Short"]
---

# Glance: The Generator Is Its Own Reranker

## 速览

Retrieval-augmented generation commonly serves separate models for reranking and answering, duplicating model residency and passage processing. We introduce Glance, which equips the generator with a lightweight decision…

## 关键点

- Retrieval-augmented generation commonly serves separate models for reranking and answering, duplicating model residency and passage processing.
- We introduce Glance, which equips the generator with a lightweight decision adapter and pointer head to score candidate passages from a shared prefill.
- Short relevance probes branch from the cached context, producing one score per passage without separately prefilling each query-passage pair.

## 全文

Retrieval-augmented generation commonly serves separate models for reranking and answering, duplicating model residency and passage processing. We introduce Glance, which equips the generator with a lightweight decision adapter and pointer head to score candidate passages from a shared prefill. Short relevance probes branch from the cached context, producing one score per passage without separately prefilling each query-passage pair. A single resident 9B backbone supports generation, general decisions and reranking, with approximately 0.1 GiB of additional memory per decision skill. After 39 minutes of MS MARCO fine-tuning, Glance achieves nDCG@10 of 0.741 on SciFact and 0.345 on NFCorpus, compared with 0.721 and 0.326 for bge-reranker-v2-m3. On 300 HotpotQA questions, its selected context improves answer F1 by 0.045 over cross-encoder-selected context under the same base generator (paired bootstrap 95% CI [0.014, 0.076]). In a scaling study, reranking quality keeps rising up to 200 candidates (43k tokens) scored from one shared prefill, where it exceeds the cross-encoder by 0.058 nDCG@10, and a probe-only activated adapter lets generation continue from the original passage cache. These initial results demonstrate generator-based reranking with one resident backbone; evaluation currently covers two retrieval datasets, one QA dataset and one backbone family.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_022256a0b411f27fc6716a1d