Long-Horizon Embodied Decision-Making via Multimodal Memory Compression

• URL: http://arxiv.org/abs/2608.01456v1
• Date: Sun, 02 Aug 2026 19:29:48 GMT
• Title: Long-Horizon Embodied Decision-Making via Multimodal Memory Compression
• Abstract summary: DunphyBench is a new benchmark for evaluating agents on long-horizon human-centered embodied decision-making.<n>MeMento is a preference-conditioned multimodal memory compressor that selectively compresses decision-relevant information based on user preferences.<n>MeMento helps VLM-driven agents improve accuracy by 7.18%, while reducing memory usage by 85.38% compared to the strongest baseline.
• Score: 71.16335993005139
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Agents are increasingly expected to act not only as task executors, but also as decision-makers on behalf of human users. This shift requires agents to accumulate evidence over long horizons, interpret implicit user preferences, and compare multiple candidates under partial observations. In this work, we propose DunphyBench, a new benchmark for evaluating agents on long-horizon human-centered embodied decision-making, where the agent must navigate through multiple embodied housing environments and make decisions that align with multi-dimensional human preferences. Unlike standard embodied reasoning tasks that often focus on procedural planning or immediate goal completion, our setting requires agents to integrate multimodal, multi-source input into coherent knowledge that supports complex reasoning across long horizon. The evaluation results reveal that there is a substantial gap between current agents and human performance. Furthermore, our diagnosis of state-of-the-art VLM-driven agents reveals that memory management is one of the bottlenecks, where raw multimodal history introduces noise that hinders decision quality. Motivated by this finding, we design MeMento, a preference-conditioned multimodal memory compressor that selectively compresses decision-relevant information from long-horizon history based on user preferences with a fixed set of memory tokens. Experiments show that MeMento helps VLM-driven agents improve accuracy by 7.18%, while reducing memory usage by 85.38% compared to the strongest baseline.

Related papers

• FinPerMA: A Theory-Informed, Event-Grounded Personalized-Memory Benchmark for LLM Agents [9.305912044747368]
  FinPerMA is an event-grounded benchmark that evaluates personalized memory against frozen longitudinal investor trajectories.<n>No full-context configuration exceeds approximately 0.47 overall accuracy or approximately 39% on multiple-choice questions.<n> Attribution analysis shows that summary-based memory often preserves factual details while losing the preference signals needed for personalization.
  arXiv  Detail & Related papers  (2026-08-04T18:00:04Z)
• VitaBench 2.0: Evaluating Personalized and Proactive Agents in Long-Term User Interactions [63.13827503828231]
  We introduce VitaBench 2.0, a benchmark for evaluating personalized and proactive agent behavior in long-term user interactions.<n>Results show that real-world personalization remains highly challenging even for state-of-the-art models.
  arXiv  Detail & Related papers  (2026-05-26T15:07:38Z)
• MINTEval: Evaluating Memory under Multi-Target Interference in Long-Horizon Agent Systems [69.06764269022925]
  We study how current memory-augmented agents perform in realistic, interference-heavy, long-horizon settings.<n> MINTEval is a benchmark featuring long, highly interconnected contexts with frequently updated information.<n>Overall, MINTEval has 15.6k question-answering pairs over long-horizon contexts averaging 138.8k tokens and extending to 1.8M tokens per instance.
  arXiv  Detail & Related papers  (2026-05-18T15:43:35Z)
• HabitatAgent: An End-to-End Multi-Agent System for Housing Consultation [8.894547553174196]
  We present HabitatAgent, the first LLM-powered multi-agent architecture for end-to-end housing consultation.<n> HabitatAgent comprises four specialized agent roles: Memory, Retrieval, Generation, and Validation.<n>We evaluate HabitatAgent on 100 real user consultation scenarios under an end-to-end correctness protocol.
  arXiv  Detail & Related papers  (2026-04-01T06:59:50Z)
• Explore with Long-term Memory: A Benchmark and Multimodal LLM-based Reinforcement Learning Framework for Embodied Exploration [52.35887679314727]
  Long-term Memory Embodied Exploration aims to unify the agent's exploratory cognition and decision-making behaviors.<n>To enhance the agent's memory recall and proactive exploration capabilities, we propose MemoryExplorer.
  arXiv  Detail & Related papers  (2026-01-11T16:23:22Z)
• Hello Again! LLM-powered Personalized Agent for Long-term Dialogue [63.65128176360345]
  We introduce a model-agnostic framework, the Long-term Dialogue Agent (LD-Agent)<n>It incorporates three independently tunable modules dedicated to event perception, persona extraction, and response generation.<n>The effectiveness, generality, and cross-domain capabilities of LD-Agent are empirically demonstrated.
  arXiv  Detail & Related papers  (2024-06-09T21:58:32Z)
• Sequential Recommender via Time-aware Attentive Memory Network [67.26862011527986]
  We propose a temporal gating methodology to improve attention mechanism and recurrent units. We also propose a Multi-hop Time-aware Attentive Memory network to integrate long-term and short-term preferences. Our approach is scalable for candidate retrieval tasks and can be viewed as a non-linear generalization of latent factorization for dot-product based Top-K recommendation.
  arXiv  Detail & Related papers  (2020-05-18T11:29:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.