Profiles2021-2026
Ziqian Zeng
South China University of Technology
19Papers
8Last author
29Unique co-authors
2First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers19
- ARGUS: Defending Against Multimodal Indirect Prompt Injection via Steering Instruction-Following BehaviorAuthor 2 of 8CVPR 2026 · 6 citations
- MixKVQ: Query-Aware Mixed-Precision KV Cache Quantization for Long-Context ReasoningAuthor 2 of 5ACL 2026 · 3 citations
- Activation-Guided Local Editing for Jailbreaking AttacksAuthor 4 of 7ACL 2026 · 1 citation
- Preserving Language Capabilities in Vision-Language Models via Representation RegulationAuthor 3 of 3ACL Findings 2026
- PrivacyRestore: Privacy-Preserving Inference in Large Language Models via Privacy Removal and RestorationAuthor 1 of 7ACL 2025 · 31 citations
- GenderAlign: An Alignment Dataset for Mitigating Gender Bias in Large Language ModelsAuthor 2 of 7ACL 2025 · 18 citations
- SEA: Low-Resource Safety Alignment for Multimodal Large Language Models via Synthetic EmbeddingsAuthor 5 of 5ACL 2025 · 5 citations
- DrFrattn: Directly Learn Adaptive Policy from Attention for Simultaneous Machine TranslationAuthor 3 of 3EMNLP 2025 · 2 citations
- Context Reasoner: Incentivizing Reasoning Capability for Contextualized Privacy and Safety Compliance via Reinforcement LearningAuthor 5 of 11EMNLP 2025 · 1 citation
- RewardDS: Privacy-Preserving Fine-Tuning for Large Language Models via Reward Driven Data SynthesisAuthor 8 of 8EMNLP 2025
- SDD: Self-Degraded Defense against Malicious Fine-tuningAuthor 4 of 4ACL 2025
- DS-AL: A Dual-Stream Analytic Learning for Exemplar-Free Class-Incremental LearningAuthor 4 of 6AAAI 2024 · 51 citations
- GACL: Exemplar-Free Generalized Analytic Continual LearningAuthor 7 of 8NeurIPS 2024 · 36 citations
- ConsistentEE: A Consistent and Hardness-Guided Early Exiting Method for Accelerating Language Models InferenceAuthor 1 of 5AAAI 2024 · 26 citations
- F-OAL: Forward-only Online Analytic Learning with Fast Training and Low Memory Footprint in Class Incremental LearningAuthor 5 of 8NeurIPS 2024 · 17 citations
- PsFuture: A Pseudo-Future-based Zero-Shot Adaptive Policy for Simultaneous Machine TranslationAuthor 3 of 3EMNLP 2024 · 1 citation
- From Ultra-Fine to Fine: Fine-tuning Ultra-Fine Entity Typing Models to Fine-grainedAuthor 2 of 2ACL 2023 · 4 citations
- Adaptive Policy with Wait-k Model for Simultaneous TranslationAuthor 6 of 7EMNLP 2023 · 2 citations
- Ultra-Fine Entity Typing with Weak Supervision from a Masked Language ModelAuthor 2 of 2ACL 2021
