12 citations · 15 across the 11 of their papers we have counts for
12 papers
The Answer-Basin Representation Hypothesis: We Are Not Probing or Steering Concepts
Manjiang Yu, Hongji Li, Zihan Wang +5
Linear probing and activation steering use linear directions to predict and control behavior-level concepts such as correctness, safety, and social bias in question answering. We c…
T-IMPACT: A Severity-Aware Benchmark for Contextual Image-Text Manipulation
Gagandeep Singh, Aaditya Yadav, Priyanka Singh
Recent advances in vision-language models and generative editing systems have made it increasingly easy to produce persuasive multimodal misinformation by altering images, text, or…
Multi-Adapter Representation Interventions via Energy Calibration
Manjiang Yu, Hongji Li, Junwei Chen +4
Representation intervention has emerged as a promising paradigm for aligning large language models toward desired behaviors without modifying model weights. Existing methods typica…
D-SECURE: Dual-Source Evidence Combination for Unified Reasoning in Misinformation Detection
Samudi Amarasinghe, Gagandeep Singh, Priyanka Singh
Multimodal misinformation increasingly mixes realistic im-age edits with fluent but misleading text, producing persuasive posts that are difficult to verify. Existing systems usual…
PIXEL: Adaptive Steering Via Position-wise Injection with eXact Estimated Levels under Subspace Calibration
Manjiang Yu, Hongji Li, Priyanka Singh +3
Reliable behavior control is central to deploying large language models (LLMs) on the web. Activation steering offers a tuning-free route to align attributes (e.g., truthfulness) t…
DGM4+: Dataset Extension for Global Scene Inconsistency
Gagandeep Singh, Samudi Amarsinghe, Priyanka Singh +1
The rapid advances in generative models have significantly lowered the barrier to producing convincing multimodal disinformation. Fabricated images and manipulated captions increas…
SGS: Segmentation-Guided Scoring for Global Scene Inconsistencies
Gagandeep Singh, Samudi Amarsinghe, Urawee Thani +3
We extend HAMMER, a state-of-the-art model for multimodal manipulation detection, to handle global scene inconsistencies such as foreground-background (FG-BG) mismatch. While HAMME…
Adaptive Token-Weighted Differential Privacy for LLMs: Not All Tokens Require Equal Protection
Manjiang Yu, Priyanka Singh, Xue Li +1
Large language models (LLMs) frequently memorize sensitive or personal information, raising significant privacy concerns. Existing variants of differential privacy stochastic gradi…
From Prediction to Explanation: Multimodal, Explainable, and Interactive Deepfake Detection Framework for Non-Expert Users
Shahroz Tariq, Simon S. Woo, Priyanka Singh +3
The proliferation of deepfake technologies poses urgent challenges and serious risks to digital integrity, particularly within critical sectors such as forensics, journalism, and t…
Random Client Selection on Contrastive Federated Learning for Tabular Data
Achmad Ginanjar, Xue Li, Priyanka Singh +1
Vertical Federated Learning (VFL) has revolutionised collaborative machine learning by enabling privacy-preserving model training across multiple parties. However, it remains vulne…
Continual Contrastive Learning on Tabular Data with Out of Distribution
Achmad Ginanjar, Xue Li, Priyanka Singh +1
Out-of-distribution (OOD) prediction remains a significant challenge in machine learning, particularly for tabular data where traditional methods often fail to generalize beyond th…
Representation Learning on Out of Distribution in Tabular Data
Achmad Ginanjar, Xue Li, Priyanka Singh +1
The open-world assumption in model development suggests that a model might lack sufficient information to adequately handle data that is entirely distinct or out of distribution (O…