ScrapMem: A Bio-inspired Framework for On-device Personalized Agent Memory via Optical Forgetting
This work addresses the critical problem of long-term memory storage for LLM agents on resource-constrained edge devices, offering a practical solution that balances performance and efficiency.
ScrapMem introduces a bio-inspired framework for on-device personalized agent memory that uses optical forgetting to reduce storage costs by up to 93% while achieving state-of-the-art performance (51.0% Joint@10) on the ATM-Bench benchmark.
Long-term personalized memory for LLM agents is challenging on resource-limited edge devices due to high storage costs and multimodal complexity. To address this, we propose ScrapMem, a framework that integrates multimodal data into "Scrapbook Page." ScrapMem introduces Optical Forgetting, an optical compression mechanism that progressively reduces the resolution of older memories, lowering storage cost while suppressing low-value details. To maintain semantic consistency, we construct an Episodic Memory Graph (EM-Graph) that organizes key events into a causal-temporal structure. Extensive experiments on the multimodal ATM-Bench showcase that ScrapMem provides three main benefits: (1) strong performance, achieving a new state-of-the-art with a 51.0% Joint@10 score; (2) high storage efficiency, reducing memory usage by up to 93% via optical forgetting; and (3) improved recall, increasing Recall@10 to 70.3% through structured aggregation. ScrapMem offers an effective and storage-efficient solution for on-device long-term memory in multimodal LLM agents.