Salvato in:
Dettagli Bibliografici
Autori principali: Lin, Jingjie, Wang, Bingbing, Wang, Zihan, Jin, Zhengda, Qiao, Weiming, Li, Jing, Xu, Ruifeng
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2606.01223
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
Sommario:
  • Despite substantial progress in long-context modeling, existing benchmarks remain confined to factual memory for explicit recall, failing to measure the reflective memory required to synthesize fragmented, multimodal cues into high-level interpretations. To address this gap, we introduce RefMem-Bench, a benchmark for reflective memory in long-horizon dialogue. RefMem-Bench contains 26K annotated QA instances with eight reflective-memory dimensions and three task formats, requiring models to move beyond surface-level retrieval and infer latent meanings from evidence distributed across interaction histories. To enhance reflective memory capability, we propose REflective Memory INDuction (REMIND), a hierarchical framework that treats reflective memory as progressive meaning construction. REMIND couples question-conditioned evidence retrieval, salience-aware grounding, and abstraction-level supervision, and uses Progressive Reflective Alignment to distill high-level reflective reasoning into the factual inference pathway. Experiments show RefMem-Bench poses a substantial challenge to current models, while REMIND consistently improves both answer accuracy and memory recall through progressive evidence perception, grounding, and abstraction.