Search papers, labs, and topics across Lattice.
This paper introduces LocalLSTC, a novel architecture designed to enhance the control of locally deployed GUI agents by organizing control information based on temporal scope. By maintaining persistent cross-step state, LocalLSTC improves the execution of short-term commitments while ensuring long-term goal alignment, addressing significant control failures identified in existing frameworks. The results show that LocalLSTC achieves a substantial performance increase, reaching 64.7% SR-100 on OSWorld and 65.3% on WindowsAgentArena, surpassing previous local benchmarks with Qwen3.6-27B.
LocalLSTC reveals that organizing control information temporally can drastically improve the performance of GUI agents, achieving over 64% success rates where previous models faltered.
Modern GUI-agent frameworks achieve strong desktop task performance with frontier API models, yet persistent control information often remains implicit in growing interaction trajectories. At each step, the planner reconstructs the active task stage, accumulated evidence, and runtime feedback before deciding the next action. This dependence becomes more pronounced under weaker local reasoning backbones. Across four representative state-of-the-art frameworks, replacing GPT-5 with Qwen3.5-9B reduces average OSWorld SR-100 from 60.9\% to 37.7\%. Trajectory annotation further identifies at least one control failure in 91.6\% of failed trajectories. To address this problem, we introduce LocalLSTC, a training-free architecture that organizes control by temporal scope, maintaining persistent cross-step state to guide short-term execution commitments. Long-Term Control maintains the active subgoal, subgoal-aligned evidence, and runtime feedback across interactions, while Short-Term Execution realizes bounded commitments for the current step. Long-to-Short Planning forms each commitment from persistent state, and Short-to-Long Control integrates execution outcomes back into that state for progress assessment, recovery, and termination. With Qwen3.6-27B, LocalLSTC reaches 64.7\% SR-100 on OSWorld and 65.3\% on WindowsAgentArena, outperforming the strongest prior local results on both benchmarks. Ablations further support contributions from mechanisms on both sides of execution. These findings identify temporal organization of control information as a distinct architectural dimension for locally deployed GUI agents.
