Real-world event detection systems rarely work with a complete list of event types — new kinds of events keep emerging. When an event type is missing from the predefined set, existing zero-shot detectors either miss the event entirely or force it into a plausible but wrong known type, silently hiding the need for discovery. We propose OGED, an ontology-guided framework that treats known-type assignment and new-type discovery as competing explanations and resolves them jointly before output. On a unified benchmark with 564 event types, OGED substantially improves both known-type detection and unknown-type discovery over strong baselines.

We study the problem of user preference drift in LLM-based recommendation and propose RAIE, a region-aware incremental editing framework. Instead of global updates or instance-level edits, RAIE introduces preference regions as structured units for localized adaptation. This design enables efficient continual learning while preserving stable preferences.

Safety-aligned LLMs frequently generate refusal responses to harmless queries due to superficial lexical similarity with malicious ones — a phenomenon known as over-refusal. Existing approaches either reduce over-refusals or preserve safety, but rarely achieve both simultaneously. We propose AdaCD, a training-free and model-agnostic adaptive contrastive decoding method that dynamically adjusts the refusal token distribution to mitigate over-refusal while maintaining or even enhancing model safety.

Large Language Models (LLMs) need to be both helpful and safe, but achieving both is a major challenge. We propose MidPO, a Mixture of Experts (MoE) framework that fine-tunes two specialized experts for safety and helpfulness and uses a dynamic routing mechanism to adaptively balance them, outperforming existing methods.

Dual-perspective, dual-granularity fairness evaluation for LLM-based job recommendation: we assess bias from both the user and the recruiter sides, at individual and group levels.

AI-generated images are more visually stunning than ever, but how do they really stack up against natural, real-world photos? We introduce D-Judge, a large-scale benchmark designed to systematically investigate and quantify the discrepancies that remain.

After careful design and development, the official website of the Knowledge Intelligence Lab (KILab) at Sun Yat-sen University is now fully upgraded and online! The new website adopts modern design concepts, providing visitors with a clearer and more intuitive browsing experience.

We are pleased to announce that Professor Ziyu Lyu officially joined the School of Cyber Science and Technology at Sun Yat-sen University in September 2023, serving as Associate Professor and Ph.D. Supervisor.