Searcharxiv⌕ Search

arXiv subjects

Yansu Wang

Publications and source records attributed to Yansu Wang.

3 recordsLinked to original sources

SAFETYEDIT: Sparse Internal Targets with Prompt-Wide Control for LLM Jailbreak Backdoors

Model-editing backdoors aim to induce harmful responses under a trigger while preserving ordinary behavior on untriggered input. Recent model-editing approaches target affirmative response prefixes through rank-one updates to a small subset of model weights. Yet when this prefix-level target is satisfied, the model may still revert to refusal or produce a coherent, apparently compliant response that does not address the request. To overcome this limitation, existing internal-representation approaches shift focus from output tokens to dense, high-dimensional compliance directions in latent space. Existing representation-level approaches rely on preservation constraints to maintain conditionality, motivating us to ask whether a smaller internal target can offer a different effectiveness-preservation trade-off. Motivated by the comparable control capability of single-coordinate interventions, we investigate a sparse refusal-related MLP coordinate as this target. We propose SAFETYEDIT, which optimizes a single refusal-related coordinate and uses KV-cache marking in early attention layers to propagate a prefix-trigger signal across subsequent instruction positions. On some checkpoints a downstream MLP reader converts this signal into control of the coordinate, while on others the propagated signal alone suffices; the method is defined by the sparse target rather than by a fixed writer-reader pair. Across four open-weight LLMs and three dataset groups, SAFETYEDIT achieves the highest triggered Guard ASR in 11 of 12 comparisons and the highest mean StrongREJECT score in all 12, relative to the baselines evaluated for each model. Crucially, comprehensive ablation and behavioral audits confirm generally minimal changes in untriggered behavior and general capabilities in the tested configurations.

cs.CL↗

Multi-Source Multi-View Graph Domain Adaptation with Hyperbolic Residual Encoding for Cross-Site MDD Identification from rs-fMRI

Cross-site identification of major depressive disorder (MDD) from resting-state functional magnetic resonance imaging (rs-fMRI) is hindered by inter-site distribution shifts and heterogeneous functional connectivity (FC) views. These views capture complementary neural relationships but exhibit distinct site biases and graph topologies, complicating alignment without sacrificing disease-relevant information or cross-view consistency. Existing studies largely treat multi-view connectome learning and cross-site adaptation separately. To the best of our knowledge, few studies have jointly modeled multiple FC views under multi-source unsupervised domain adaptation for cross-site rs-fMRI-based MDD classification. We construct Pearson correlation, sparse representation, and Granger causality graphs, each encoded by a view-specific graph attention network. Dual-stream adaptive fusion explicitly integrates pairwise cross-view interactions, followed by lightweight hyperbolic residual encoding for curvature-aware representation refinement. Class-wise Cauchy--Schwarz alignment reduces inter-source and source-target discrepancies, complemented by adversarial learning, information maximization, and confidence-aware pseudo-labeling. Across seven unlabeled target domains, our framework achieves 73.60% mean accuracy and 71.90% AUC, demonstrating effective generalization under heterogeneous acquisition conditions. These results highlight the effectiveness of unified heterogeneous-view modeling, curvature-aware refinement, and multi-source domain adaptation for cross-site MDD identification.The source code is at https://github.com/OPUS-Lightphenexx/MM-HyperGDA

cs.CV↗

Partial domain adaptation enables cross domain cell type annotation between scRNA-seq and snRNA-seq

Accurate cell type annotation across datasets is a key challenge in single-cell analysis. snRNA-seq enables profiling of frozen or difficult-to-dissociate tissues, complementing scRNA-seq by capturing fragile or rare cell types. However, cross-annotation between these two datasets remains largely unexplored, as existing methods treat them independently. We introduce ScNucAdapt, a method designed for cross-annotation between paired and unpaired scRNA-seq and snRNA-seq datasets. To address distributional and cell composition differences, ScNucAdapt employs partial domain adaptation. Experiments across both unpaired and paired scRNA-seq and snRNA-seq show that ScNucAdapt achieves robust and accurate cell type annotation, outperforming existing approaches. Therefore, ScNucAdapt provides a practical framework for the cross-domain cell type annotation between scRNA-seq and snRNA seq data.

q-bio.GN↗