Zhichen Li | Searcharxiv

arXiv2026

DART-VLN: Test-Time Memory Decay and Anti-Loop Regularization for Discrete Vision-Language Navigation

Memory-based agents for discrete vision-language navigation (VLN) operate under partial observability and can exhibit systematic inference-time failures even with strong pretrained backbones. We focus on two recurring problems: stale historical evidence during memory readout and inefficient local backtracking during action selection. We present DART-VLN, a training-free inference-time framework that combines Test-Time Memory Decay, which reweights stale and redundant memory slots without modifying their stored content, with Anti-Loop Regularization, a lightweight next-hop penalty that discourages immediate reversals. DART-VLN introduces no learnable parameters and leaves the navigation backbone unchanged. Experiments on R2R and REVERIE show that memory decay consistently preserves or improves task performance while reducing runtime. Adding anti-loop regularization further shortens trajectories, reduces local backtracking, and achieves the best overall balance between navigation quality and efficiency among the evaluated GridMM variants. These results indicate that lightweight inference-time control can improve the reliability and efficiency of memory-based discrete VLN without retraining.

cs.RO↗

arXiv2023

Multi-label Learning from Privacy-Label

Multi-abel Learning (MLL) often involves the assignment of multiple relevant labels to each instance, which can lead to the leakage of sensitive information (such as smoking, diseases, etc.) about the instances. However, existing MLL suffer from failures in protection for sensitive information. In this paper, we propose a novel setting named Multi-Label Learning from Privacy-Label (MLLPL), which Concealing Labels via Privacy-Label Unit (CLPLU). Specifically, during the labeling phase, each privacy-label is randomly combined with a non-privacy label to form a Privacy-Label Unit (PLU). If any label within a PLU is positive, the unit is labeled as positive; otherwise, it is labeled negative, as shown in Figure 1. PLU ensures that only non-privacy labels are appear in the label set, while the privacy-labels remain concealed. Moreover, we further propose a Privacy-Label Unit Loss (PLUL) to learn the optimal classifier by minimizing the empirical risk of PLU. Experimental results on multiple benchmark datasets demonstrate the effectiveness and superiority of the proposed method.

cs.LG↗