# Glance: The Generator Is Its Own Reranker

**Authors:** Adilet Metinov, Chingiz Arziev  
**Lab:** Airun Intelligence Labs · **Published:** 29 Oct 2026 · **Upvotes:** 12  
**Links:** [arXiv](https://arxiv.org/abs/2610.40454) · [PDF](https://arxiv.org/pdf/2610.40454) · [This page](https://omen-bit.github.io/ai-paper-daily/paper/2610.40454.html)

## Key numbers

- **95%**: paired bootstrap 95% CI [0.014
- **300**: On 300 HotpotQA questions

## Abstract

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.

---

Notes written by AI from the paper, numbers copied from the abstract. Check the paper before you rely on them. From AI Papers Daily.