ScopeMem: Latent Memory That Learns Where to Revise Frozen GUI Agents
Latent memory is gaining ground as a lightweight history supplement for GUI agents.
ProofPaper ↗
Key points
- It supplies history to a frozen vision-language model via prefix vectors trained with step-wise cross-entropy (CE).
- The prefix learns to alter the model's default actions.
- We therefore propose ScopeMem, a latent-memory method for scoped revision.
- On AndroidControl, GUI-Odyssey, and zero-shot AITW, ScopeMem outperforms all evaluated prompting and latent-memory baselines, enabling a frozen Qwen3-VL-8B to match or surpass GUI-specialized agents and Qwen3-VL-32B.
Sources (1)
- [1]ScopeMem: Latent Memory That Learns Where to Revise Frozen GUI AgentsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 19, 02:38 PM
Latent memory is gaining ground as a lightweight history supplement for GUI agents.
It supplies history to a frozen vision-language model via prefix vectors trained with step-wise cross-entropy (CE).
Extractive summary: sentences quoted from the sources.
Before this
- Oct 19, 2026MiniMax M3.1 Flash shows China's AI price race has no floor in sight
- Oct 19, 2026LEAPS: Long-context Evaluation of Adaptive Pattern Learning and State Tracking
- Oct 19, 2026PVBench: Quantifying Physical Plausibility in Video Generation with Object-Level Physical Measurements
- Oct 19, 2026Remembering Has a Price: Learning Memory Worth under Resource Constraints
- Oct 19, 2026VerifierBench: Benchmarking the Generation of Executable Verifiers for LLM Agents
- Oct 18, 2026Editing as Language: Teaching LLM Agents a Grammar for Native Context Editing