10 papers
Narrative Paradoxes in LLM Inference: Shaping Reasoning Trajectories toward Misalignment
Haoming Yang, Ke Ma, Yangbangyan Jiang +6
Explicit reasoning in large language models (LLMs) is often associated with reliability, yet it can still pose risks when a model infers harmful goals under a benign context. We id…
New bounds for off-diagonal hypergraph Ramsey numbers
David Conlon, Xiaoyu He, Jiaxi Nie +1
We study a conjecture of Conlon, Fox, Gunby, He, Mubayi, Suk, Verstraëte, and Yu, which characterizes the -graphs for which the off-diagonal Ramsey number is polyno…
From Context-Aware to Conflict-Aware: Generalizing Contrastive Decoding for Knowledge Conflict in LLMs
Runze Jiang, Taiqiang Wu, Yan Wang +2
When large language models generate from retrieved or augmented contexts, conflicts between external context and parametric priors remain a central reliability bottleneck. Existing…
How LoRA Remembers? A Parametric Memory Law for LLM Finetuning
Ziwen Xu, Haiwen Hong, Linsong Yu +4
Large Language Models (LLMs) must continuously learn and update knowledge to remain effective in dynamic real-world environments. While Low-Rank Adaptation (LoRA) is widely used fo…
Robust and Generalizable Safety Steering for Text-to-Image Diffusion Transformers
Zihao Xue, Yan Wang, Zhen Bi +7
Diffusion Transformers have become a powerful backbone for text-to-image generation, but their layered and cross-modal generation process makes safety control fundamentally differe…
Make LLM Learn to Synthesize from Streaming Experiences through Feedback
Zhenlin Hu, Yan Wang, Zhen Bi +7
Large language models (LLMs) have been widely adopted for synthetic data generation, significantly reducing annotation costs. However, most existing studies treat synthesis as a se…
The Bridge-Garden Dilemma in LLM Distillation: Why Mixing Hard and Soft Labels Works
Guanghui Wang, Kaiwen Lv Kacuila, Zhiyong Yang +5
Knowledge distillation (KD) transfers knowledge from a large teacher model to a smaller student. In language modeling, the student is trained either on tokens sampled from the teac…
Localize and Neutralize: Gradient-guided Token Suppression against Visual Prompt Injection Attack
Dongpeng Zhang, Ke Ma, Yangbangyan Jiang +4
Adversarial images pose a severe security threat to multimodal large language models through prompt injection. Existing defenses largely lack a principled understanding of the unde…
Making Training-Free Diffusion Segmentors Scale with the Generative Power
Benyuan Meng, Qianqian Xu, Zitai Wang +3
As powerful generative models, text-to-image diffusion models have recently been explored for discriminative tasks. A line of research focuses on adapting a pre-trained diffusion m…
General Phrase Debiaser: Debiasing Masked Language Models at a Multi-Token Level
Bingkang Shi, Xiaodan Zhang, Dehan Kong +4
The social biases and unwelcome stereotypes revealed by pretrained language models are becoming obstacles to their application. Compared to numerous debiasing methods targeting wor…