5 papers
RuleSafe-VL: Evaluating Rule-Conditioned Decision Reasoning in Vision-Language Content Moderation
Zhifeng Lu, Dianyuan Wang, Yuhu Shang +1
Platform content moderation applies explicit policy rules and context-dependent conditions to decide whether user content is allowed, restricted, or removed. A correct moderation o…
S-Attention:Attention-Aligned Endogenous Retrieval for Memory-Bounded Long-Context Inference
Qingsen Ma, Dianyun Wang, Yaoye Wang +7
Large language models are increasingly applied to multi-document and long-form inputs, yet long-context inference remains memory- and noise-inefficient. Key-value (KV) caching scal…
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…
CIP: A Plug-and-Play Causal Prompting Framework for Mitigating Hallucinations under Long-Context Noise
Qingsen Ma, Dianyun Wang, Ran Jing +2
Large language models often hallucinate when processing long and noisy retrieval contexts because they rely on spurious correlations rather than genuine causal relationships. We pr…
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…