analiza aiqbench
This research introduces CUE-Mem, a new benchmark designed to evaluate how AI agents recall long-term memories from implicit multimodal cues like background objects and ambient sounds rather than just explicit text. The study identifies that current systems struggle significantly with retrieving these subtle signals, revealing bottlenecks in both textualized memory approaches and native multimodal indexing that introduce retrieval noise. For the AI community, this work highlights the critical need for architectures that can efficiently preserve and access peripheral sensory data without incurring prohibitive token costs or performance degradation.
CUE-Mem: Benchmarking Long-Term User Memory via Implicit Cues in Multimodal Conversations
arXiv:2609.32574v1 Announce Type: cross Abstract: Long-term memory is essential for multimodal agents that interact with users across sustained conversations. However, user memories are not always explicitly stated: they may also be implied by recurring background objects in images, ambient sounds in audio, or other peripheral multimodal cues. Existing benchmarks largely focus on text-only memory or explicit multimodal evidence, leaving implicit multimodal cues underexplored. We introduce CUE-Mem, a text-image-audio benchmark for evaluating long-term user memory from implicit cues. CUE-Mem contains 2,674 questions across explicit and implicit evidence settings and covers four tasks: Entity Recall, Long Pattern, Personalized Recommendation, and Answer Refusal. Across textualized memory systems, implicit performance remains far below oracle evidence, locating the main bottleneck in preserving and retrieving subtle cues rather than question answerability. Increasing caption detail recovers more of this evidence, but brings uneven gains and rapidly growing token costs, motivating native multimodal access. Yet native access does not uniformly resolve the bottleneck: evidence use depends strongly on the backbone, while multimodal indexing introduces substantial retrieval noise. CUE-Mem provides a testbed for memory systems that selectively retain, retrieve, and use subtle multimodal evidence.
Abstract:Long-term memory is essential for multimodal agents that interact with users across sustained conversations. However, user memories are not always explicitly stated: they may also be implied by recurring background objects in images, ambient sounds in audio, or other peripheral multimodal cues. Existing benchmarks largely focus on text-only memory or explicit multimodal evidence, leaving implicit multimodal cues underexplored. We introduce CUE-Mem, a text-image-audio benchmark for evaluating long-term user memory from implicit cues. CUE-Mem contains 2,674 questions across explicit and implicit evidence settings and covers four tasks: Entity Recall, Long Pattern, Personalized Recommendation, and Answer Refusal. Across textualized memory systems, implicit performance remains far below oracle evidence, locating the main bottleneck in preserving and retrieving subtle cues rather than question answerability. Increasing caption detail recovers more of this evidence, but brings uneven gains and rapidly growing token costs, motivating native multimodal access. Yet native access does not uniformly resolve the bottleneck: evidence use depends strongly on the backbone, while multimodal indexing introduces substantial retrieval noise. CUE-Mem provides a testbed for memory systems that selectively retain, retrieve, and use subtle multimodal evidence.
28 pages. Submitted to AAAI 2027. Code: this https URL. Data: this https URL
Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
arXiv:2609.32574 [cs.AI]
(or arXiv:2609.32574v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2609.32574 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 292 cuvinte
Subiecte conexe
Six Principles of Voice Design: Designing for the Ear and Encoding It into Agent Script
Salesforce Blog · 8 oct. 2026
MobileMem: Learning from a Year of Mobile Experiences
arXiv · 17 aug. 2026
[AINews] OpenAI shuts off Cursor
Latent Space · 29 aug. 2026
Keep It InMind: Benchmarking the Implicit-Association Blind Spot in Agent Memory
arXiv · 28 iul. 2026