6 papers
Dynamics Reveals Structure: Challenging the Linear Propagation Assumption
Hoyeon Chang, Bálint Mucsányi, Seong Joon Oh
Neural networks adapt through first-order parameter updates, yet it remains unclear whether such updates preserve logical coherence. We investigate the geometric limits of the Line…
Characterizing Pattern Matching and Its Limits on Compositional Task Structures
Hoyeon Chang, Jinho Park, Hanseul Cho +7
Despite impressive capabilities, LLMs' successes often rely on pattern-matching behaviors, yet these are also linked to OOD generalization failures in compositional tasks. However,…
Dynamic VLM-Guided Negative Prompting for Diffusion Models
Hoyeon Chang, Seungjin Kim, Yoonseok Choi
We propose a novel approach for dynamic negative prompting in diffusion models that leverages Vision-Language Models (VLMs) to adaptively generate negative prompts during the denoi…
Latent Reasoning via Sentence Embedding Prediction
Hyeonbin Hwang, Byeongguk Jeon, Seungone Kim +7
Autoregressive language models (LMs) generate one token at a time, yet human reasoning operates over higher-level abstractions - sentences, propositions, and concepts. This contras…
How Does Vision-Language Adaptation Impact the Safety of Vision Language Models?
Seongyun Lee, Geewook Kim, Jiyeon Kim +4
Vision-Language adaptation (VL adaptation) transforms Large Language Models (LLMs) into Large Vision-Language Models (LVLMs) for multimodal tasks, but this process often compromise…
How Do Large Language Models Acquire Factual Knowledge During Pretraining?
Hoyeon Chang, Jinho Park, Seonghyeon Ye +4
Despite the recent observation that large language models (LLMs) can store substantial factual knowledge, there is a limited understanding of the mechanisms of how they acquire fac…