11 papers
Omni-RRM: Advancing Omni Reward Modeling via Automatic Rubric-Grounded Preference Synthesis
Zicheng Kong, Dehua Ma, Zhenbo Xu +9
Multimodal large language models (MLLMs) struggle with alignment due to the limitations of existing reward models (RMs), which are predominantly vision-centric, dependent on costly…
From Recognition to Reasoning: Advancing Multimodal Harmful Meme Detection via Chain-of-Thought Alignment
Hexiang Gu, Qifan Yu, Yuan Liu +4
As a multimodal communication medium that integrates images and text, memes often convey implicit harmful content through metaphors, satire, and humor, making harmful meme detectio…
Beyond Face Swapping: A Diffusion-Based Digital Human Benchmark for Multimodal Deepfake Detection
Jiaxin Liu, Jia Wang, Saihui Hou +5
In recent years, the explosive advancement of deepfake technology has posed a critical and escalating threat to public security: diffusion-based digital human generation. Unlike tr…
Interpretable Safety Alignment via SAE-Constructed Low-Rank Subspace Adaptation
Dianyun Wang, Qingsen Ma, Yuhu Shang +5
Safety alignment -- training large language models (LLMs) to refuse harmful requests while remaining helpful -- is critical for responsible deployment. Prior work established that…
Unlocking the Address Book: Dissecting the Sparse Semantic Structure of LLM Key-Value Caches via Sparse Autoencoders
Qingsen Ma, Dianyun Wang, Jiaming Lyu +8
The Key-Value (KV) cache is the primary memory bottleneck in long-context Large Language Models, yet it is typically treated as an opaque numerical tensor. In this work, we propose…
Select-Then-Decompose: From Empirical Analysis to Adaptive Selection Strategy for Task Decomposition in Large Language Models
Shuodi Liu, Yingzhuo Liu, Zi Wang +4
Large language models (LLMs) have demonstrated remarkable reasoning and planning capabilities, driving extensive research into task decomposition. Existing task decomposition metho…