DMV-Bench: Diagnosing Long-Horizon Multimodal Agents' Visual Memory with Incidental Cue Injection

Researchers introduced DMV-Bench, the first interactive benchmark for multimodal-agent visual memory, testing agents' ability to recall visual cues from product images in a simulated e-commerce environment. Their proposed DualMem architecture, which maintains parallel visual and verbal codes, outperformed existing systems across multiple chain lengths on Gemini 2.5 Flash and Qwen2.5-VL-7B. The benchmark addresses a gap in evaluating visual memory in interactive settings, where text-based memory has been the primary focus.

arXiv:2606.27499v1 Announce Type: new Abstract: Research on agent memory has matured rapidly, but almost entirely on the text side: few existing benchmarks ask, in an interactive environment, when an agent genuinely needs to remember what it saw rather than what it could write down. We introduce DMV-Bench Code: https://github.com/yyyujintang/DMV-Bench , the first interactive benchmark for multimodal-agent visual memory. DMV-Bench is built on a controlled home-furnishing e-commerce catalogue of 1,000 product variants in which a text-leakage contract keeps the discriminative signal of each task in the pixels alone. Across a chain of autonomous shopping sessions, every visited product image carries a unique, pre-rendered incidental cue, and the agent is later asked to recall a particular cued product and navigate to its URL. Inspired by dual-coding theory, we propose DualMem, a memory architecture that maintains a visual and a verbal code in parallel. On DMV-Bench, DualMem outperforms a caption baseline and three recent multimodal agent-memory systems at every chain length J in {5, 10, 15, 50} on both Gemini 2.5 Flash and Qwen2.5-VL-7B, with the lead surviving controls for memory-bank size and encoding-position bias, and an asymmetric dual-coding regime in which vision carries the cue end-to-end while the verbal channel plays a smaller query-grounding role.