Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks30 from public data
- EZH2 phosphorylation by JAK3 mediates a switch to noncanonical function in natural killer/T-cell lymphoma137
Pharmacological inhibition of JAK3 activity may provide a promising treatment option for NKTL through the novel mechanism of suppressing noncanonical EZH2 activity, which upregulates a set of genes involved in DNA replication, cell cycle, biosynthesis, stemness, and invasiveness.
- Towards Label-Only Membership Inference Attack against Pre-trained Large Language Models54
A label-only membership inference attack based on token-level semantic similarity to approximate output probabilities and subsequently calculate the perplexity, which performs better than the extensions of existing label-only attacks against personalized LLMs and even on par with other advanced logit-based attacks across all metrics on five prevalent open-source LLMs.
- Two-photon fluorescent probe for cadmium imaging in cells42
The TPCd probe was pH-independent in the biologically relevant pH range and non-toxic to living cells at reasonable concentration levels, warranting its in vivo applications.
- Nearest is Not Dearest: Towards Practical Defense Against Quantization-Conditioned Backdoor Attacks40
This paper conducts the first in-depth analysis of QCBs and proposes Error-guided Flipped Rounding with Activation Preservation (EFRAP), an effective and practical defense against QCBs that flips the rounding strategies of neurons crucial for back-door effects but with minimal impact on clean accuracy.
- Towards Reliable Verification of Unauthorized Data Usage in Personalized Text-to-Image Diffusion Models34
SIREN is introduced, a novel methodology to proactively trace unauthorized data usage in black-box personalized text-to-image diffusion models that optimizes the coating in a delicate way to be recognized by the model as a feature relevant to the personalization task, thus significantly improving its learnability.
- Is Difficulty Calibration All We Need? Towards More Practical Membership Inference Attacks28
The observations and analysis challenge the current de facto paradigm of difficulty calibration in high-precision inference, encouraging greater attention to the persistent risks posed by MIAs in more practical scenarios.
- Free Fine-tuning: A Plug-and-Play Watermarking Scheme for Deep Neural Networks24
This paper proposes a plug-and-play watermarking scheme for DNN models by injecting an independent proprietary model into the target model to serve the watermark embedding and ownership verification, thus the fidelity is well preserved and scalable to challenging real tasks.
- What can Discriminator do? Towards Box-free Ownership Verification of Generative Adversarial Networks19
A novel IP protection scheme for GANs where ownership verification can be done by checking outputs only, without choosing the inputs (i.e., box-free setting), that makes use of the unexploited potential of the discriminator to learn a hypersphere that captures the unique distribution learned by the paired generator.
- PSY-1, a Taxus chinensis var. mairei Extract, Inhibits Cancer Cell Metastasis by Interfering with MMPs18
The results illustrated that PSY-1 effectively suppressed the migration and invasion ability of the melanoma cancer cell line B16-F10, caused down-regulation of MMP-2 and M MP-9, and that the NF-κB pathway was involved in the anti-metastatic effects imposed by PSy-1.
- DATABench: Evaluating Dataset Auditing in Deep Learning from an Adversarial Perspective15
A comprehensive study evaluating dataset auditing from an adversarial perspective, including a new benchmark, DATABench, comprising 17 evasion attacks, 5 forgery attacks, and 9 representative auditing methods that reveal that none of the evaluated auditing methods are sufficiently robust or distinctive under adversarial settings.
- TraceEvader: Making DeepFakes More Untraceable via Evading the Forgery Model Attribution15
This work proposes a novel training-free evasion attack, TraceEvader, which injects a universal imitated traces learned from wild DeepFakes into the high-frequency component and introduces adversarial blur into the domain of the low-frequency component, where the added distortion confuses the extraction of certain traces for model attribution.
- Invisible Backdoor Attack against Self-supervised Learning12
This paper proposes an imperceptible and effective backdoor attack against self-supervised models using optimized triggers disentangled with the augmented transformation in the SSL, while remaining imperceptible to human vision.
- Protection of CAPE-pNO2 Against Chronic Myocardial Ischemia by the TGF-Β1/Galectin-3 Pathway In Vivo and In Vitro11
It is indicated that CAPE-pNO2 may ameliorate CMI by suppressing fibrosis, inflammation, and apoptosis via the TGF-β1/Gal-3 pathway in vivo and in vitro.
- 11
- Targeting the ubiquitin pathway in lymphoid malignancies10
The altered signalling mechanisms mediated by the aberrant expression of cancer-associated E2s/E3s and deubiquitinating enzymes, which result in the hyperactivation of oncoproteins or the frequently allied downregulation of tumour suppressors are reviewed.
- Dual-level Interaction for Domain Adaptive Semantic Segmentation10
This paper proposes a dual-level interaction for domain adaptation for domain adaptation (DIDA) in semantic segmentation, which outperforms the state-of-the-art by a notable margin, especially on confusing and long-tailed classes.
- DREAM: Scalable Red Teaming for Text-to-Image Generative Systems via Distribution Modeling9
DREAM is proposed, a scalable red teaming framework to automatically uncover diverse problematic prompts from a given T2I system, and GC-SPSA, an efficient optimization algorithm that provides stable gradient estimates through the long and potentially non-differentiable T2I pipeline.
- Janus Kinase Signaling: Oncogenic Criminal of Lymphoid Cancers9
This work discusses mutations and fusions leading to JAK deregulations, how upstream nodes drive JAK expression, how classical JAK/STAT pathways are represented in lymphoid malignancies and the noncanonical and nuclear role of JAKs.
- Revisiting Data Auditing in Large Vision-Language Models7
This study presents a systematic view of the limits and opportunities of MI for VLMs, providing guidance for future efforts in trustworthy data auditing.
- Transfer Learning of Real Image Features with Soft Contrastive Loss for Fake Image Detection6
This paper introduces a self-supervised feature mapping process for natural trace extraction and develops a transfer learning based on soft contrastive loss to bring them closer to real images and further away from fake ones, underscoring its practicality for real-world application deployment.
- EZH2 K63-polyubiquitination affecting migration in extranodal natural killer/T-cell lymphoma5
It is found out dexamethasone treatment manipulates TRIP12-EZH2 signaling and may represent a novel therapeutic strategy against ENKTL metastasis, which promotes ENKTL cell migration and could be targeted by dexamethasone treatment.
- 4
- Optimized Design Method for Satellite Constellation Configuration Based on Real-time Coverage Area Evaluation3
An optimized design of satellite constellation configuration for full coverage of large-scale regional imaging is proposed by using an improved simulated annealing algorithm combined with the real-time coverage evaluation method of hexagonal discretization.
- External Data Extraction Attacks Against Retrieval-Augmented Large Language Models2
The first comprehensive study to formalize EDEAs against retrieval-augmented LLMs (RA-LLMs) and develop Secret, a Scalable and EffeCtive exteRnal data Extraction aTtack, which significantly outperforms previous attacks, and is highly effective against all 16 tested RAG instances.
- 2
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-11. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it. You will be asked to sign in.