7 papers
TriAlign: Towards Universal Truth Consistency in Personalized LLM Alignment
Thi-Nhung Nguyen, Linhao Luo, Rollin Omari +3
Personalized large language models adapt responses to users' preferences and social attributes, but can introduce substantial universal truth inconsistencies across social groups,…
MATO: Multi-objective Personalized Alignment with Test-time Optimization for Large Language Models
Linhao Luo, Thuy-Trang Vu, Van-Anh Nguyen +3
Aligning large language models (LLMs) with diverse and multifaceted user preferences is a fundamental challenge in personalized AI systems. Existing multi-objective alignment metho…
Adaptive Subspace Projection for Generative Personalization
Van-Anh Nguyen, Anh Tuan Bui, Tamas Abraham +5
Generative personalization often suffers from the semantic collapsing problem (SCP), where a learned personalized concept overpowers the rest of the text prompt, causing the model…
MCBench: A Multicontext Safety Assessment Benchmark for Omni Large Language Models
Manh Luong, Tamas Abraham, Junae Kim +6
Existing multimodal safety benchmarks focus solely on visual inputs and cannot assess Omni Large Language Models (LLMs) that process vision, audio, and text. We introduce MCBench,…
Mitigating Semantic Collapse in Generative Personalization with Test-Time Embedding Adjustment
Anh Bui, Trang Vu, Trung Le +5
In this paper, we investigate the semantic collapsing problem in generative personalization, an under-explored topic where the learned visual concept () gradually shifts from it…
Fantastic Targets for Concept Erasure in Diffusion Models and Where To Find Them
Anh Bui, Trang Vu, Long Vuong +5
Concept erasure has emerged as a promising technique for mitigating the risk of harmful content generation in diffusion models by selectively unlearning undesirable concepts. The c…