paper-with-me

홈 › Papers

STAMP: Selective Task-Aware Mechanism for Text Privacy

2026-03-12 · Fengwei Tian, Payel Bhattacharjee, Heidi Hanson, Geoffrey D. Rubin, Joseph Y. Lo, Ravi Tandon arxiv

We present STAMP (Selective Task-Aware Mechanism for Text Privacy), a new framework for task-aware text privatization that achieves an improved privacy-utility trade-off. STAMP selectively allocates privacy budgets across tokens by jointly considering (i) each token's importance to the downstream task (as measured via a task- or query-specific representation), and (ii) its privacy sensitivity (e.g., names, dates, identifiers). This token-level partitioning enables fine-grained, group-wise control over the level of noise applied to different parts of the input, balancing privacy protection with task relevance. To privatize individual token embeddings, we introduce the polar mechanism, which perturbs only the direction of embeddings on the unit sphere while preserving their magnitude. Decoding is performed via cosine nearest-neighbor search, aligning the perturbation geometry with the decoding geometry. Unlike isotropic noise mechanisms, the polar mechanism maintains semantic neighborhoods in the embedding space and better preserves downstream utility. Experimental evaluations on SQuAD, Yelp, and AG News datasets demonstrate that STAMP, when combined with the normalized polar mechanism, consistently achieves superior privacy-utility trade-offs across varying per-token privacy budgets.

📄 PDF Abstract BibTeX arXiv:2603.12237

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

TIDES: Implicit Time-Awareness in Selective State Space Models

2026-05-10 · Taylan Soydan, Miguel A. Bessa, Dirk Mohr, Rui Barreira arxiv

Selective state space models (SSMs), such as Mamba, achieve strong per-token expressivity by making the time discretization step $\TildeΔ$ a learned function of the input. However, in doing so, $\TildeΔ$ ceases to repres…

Fusing Pixels and Genes: Spatially-Aware Learning in Computational Pathology

2026-02-15 · Minghao Han, Dingkang Yang, Linhao Qu, Zizhi Chen 외 arxiv

Recent years have witnessed remarkable progress in multimodal learning within computational pathology. Existing models primarily rely on vision and language modalities; however, language alone lacks molecular specificity…

Representation Learning

SyncFed: Time-Aware Federated Learning through Explicit Timestamping and Synchronization

2025-06-11 · Baran Can Gül, Stefanos Tziampazis, Nasser Jazdi, Michael Weyrich

As Federated Learning (FL) expands to larger and more distributed environments, consistency in training is challenged by network-induced delays, clock unsynchronicity, and variability in client updates. This combination …

Federated Learning

KOSS: Kalman-Optimal Selective State Spaces for Long-Term Sequence Modeling

2025-12-18 · Lei Wang, Xin Tan, Mingwei Wang, Ying Zhang arxiv

Recent selective state space models (SSMs), such as Mamba and Mamba-2, have demonstrated strong performance in sequence modeling owing to input-dependent selection mechanisms. However, these mechanisms lack theoretical g…

Self-Attention Enhanced Selective Gate with Entity-Aware Embedding for Distantly Supervised Relation Extraction

2019-11-27 · Yang Li, Guodong Long, Tao Shen, Tianyi Zhou 외

Distantly supervised relation extraction intrinsically suffers from noisy labels due to the strong assumption of distant supervision. Most prior works adopt a selective attention mechanism over sentences in a bag to deno…

Entity EmbeddingsRelationRelation ClassificationRelation Extraction+2