🧑‍🔬 Sirui Han

🏛️ Affiliations

🎓Hong Kong University of Science and TechnologyHong Kong · China2023–2026 · 123 papersEmployment · Assistant Professorcurrent🎓Chinese University of Hong KongHong Kong · China2016–2026 · 3 papers🎓Hong Kong Polytechnic UniversityHong Kong · China2021–2023 · 3 papers

From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 1 more institution on single bylines are not listed.

🧭 Fields, topics and concepts

Of its works in our library, with the number of works on each.

📅 Works by year

2020s

📄 Papers

The most cited first (OpenAlex and our library).

1. PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference (2025 · 83 citations)
   Safety alignment of large language models depends on preference data and fine-grained harm labels, which are expensive to collect at scale.
2. FinMME: Benchmark Dataset for Financial Multi-Modal Reasoning Evaluation (2025 · 19 citations)
   FinMME addresses the lack of specialised benchmarks for testing multimodal large language models (MLLMs) on financial material, where charts are data-dense and interpretation requires expert knowledge.
3. PrivaCI-Bench: Evaluating Privacy with Contextual Integrity and Legal Compliance (2025 · 17 citations)
   Most privacy benchmarks for LLMs reduce privacy to spotting personally identifiable information. That ignores whether a particular data flow is appropriate in its setting. PrivaCI-Bench addresses this gap using Contextual Integrity (CI) theory. CI describes each flow by five parameters: sender, recipient, subject…
4. LegalReasoner: Step-wised Verification-Correction for Legal Judgment Reasoning (2025 · 16 citations)
   Legal judgment prediction (LJP) asks a model to produce a court-like ruling from a plaintiff's claim and the case facts.
5. Context Reasoner: Incentivizing Reasoning Capability for Contextualized Privacy and Safety Compliance via Reinforcement Learning (2025 · 7 citations)
   LLMs deployed as tool-using agents raise safety and privacy risks. Typical defenses detect sensitive patterns instead of reasoning about the situation, and they rarely reflect actual legal standards. The authors reframe safety and privacy as compliance questions grounded in Contextual Integrity (CI) theory. Under CI…
6. Bit-by-Bit: Progressive QAT Strategy with Outlier Channel Splitting for Stable Low-Bit LLMs (2026 · 3 citations)
   Quantization-aware training (QAT) of LLMs at extreme precisions, such as 2-bit weights and activations, often breaks down.
7. Learning While Staying Curious: Entropy-Preserving Supervised Fine-Tuning via Adaptive Self-Distillation for Large Reasoning Models (2026 · 3 citations)
   Large reasoning models are usually post-trained by supervised fine-tuning followed by reinforcement learning (SFT-then-RL).
8. Perception, Understanding and Reasoning: A Multimodal Benchmark for Video Fake News Detection (2026 · 3 citations)
   Most benchmarks for video fake news detection (VFND) score only the final real/fake label. They cannot show whether a multimodal large language model (MLLM) actually saw the right evidence, understood the story, or reasoned soundly. The authors address this with POVFNDB, a process-oriented benchmark. It is built on…
9. SafeMT: Multi-turn Safety for Multimodal Language Models (2026 · 3 citations)
   SafeMT targets a gap in the safety evaluation of multimodal large language models (MLLMs). Existing benchmarks mostly test single image-plus-prompt inputs, but real users talk to these models over many turns, and harmful intent can be spread across a conversation. The authors combined red-team queries from earlier…
10. InsightEval: An Expert-Curated Benchmark for Assessing Insight Discovery in LLM-Driven Data Agents (2026 · 2 citations)
    LLM-based data agents are increasingly used to extract insights from tables, but few benchmarks measure this ability well.

All works

14 works of theirs in our library, the most cited first; 10 with a page here, the others link their original until we summarise them.

1. PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human PreferenceAnnual Meeting of the Association for Computational Linguistics (ACL)Jiaming Ji, Donghai Hong, Borong Zhang, et al.2025 · 83 citations
2. FinMME: Benchmark Dataset for Financial Multi-Modal Reasoning EvaluationAnnual Meeting of the Association for Computational Linguistics (ACL)Junyu Luo, Zhizhuo Kou, Liming Yang, et al.2025 · 19 citations
3. PrivaCI-Bench: Evaluating Privacy with Contextual Integrity and Legal ComplianceAnnual Meeting of the Association for Computational Linguistics (ACL)Haoran Li, Wenbin Hu, Huihao Jing, et al.2025 · 17 citations
4. LegalReasoner: Step-wised Verification-Correction for Legal Judgment ReasoningAnnual Meeting of the Association for Computational Linguistics (ACL)Weijie Shi, Han Zhu, Jiaming Ji, et al.2025 · 16 citations
5. Context Reasoner: Incentivizing Reasoning Capability for Contextualized Privacy and Safety Compliance via Reinforcement LearningConference on Empirical Methods in Natural Language Processing (EMNLP)Wenbin Hu, Haoran Li, Huihao Jing, et al.2025 · 7 citations
6. LRAS: Advanced Legal Reasoning with Agentic Search ↗arXiv:2601.07296 · no summary yetYujin Zhou; Chuxue Cao; Jinluan Yang; Lijun Wu; Conghui He; Sirui Han;, Yike Guo2026 · 4 citations
7. Bit-by-Bit: Progressive QAT Strategy with Outlier Channel Splitting for Stable Low-Bit LLMsAnnual Meeting of the Association for Computational Linguistics (ACL)Binxing Xu, Hao Gu, Lujun Li, et al.2026 · 3 citations
8. Learning While Staying Curious: Entropy-Preserving Supervised Fine-Tuning via Adaptive Self-Distillation for Large Reasoning ModelsAnnual Meeting of the Association for Computational Linguistics (ACL)Hao Wang, Hao Gu, Hongming Piao, et al.2026 · 3 citations
9. Perception, Understanding and Reasoning: A Multimodal Benchmark for Video Fake News DetectionAnnual Meeting of the Association for Computational Linguistics (ACL)Cui Yakun, Peng Qi, Fushuo Huo, et al.2026 · 3 citations
10. SafeMT: Multi-turn Safety for Multimodal Language ModelsAnnual Meeting of the Association for Computational Linguistics (ACL)Han Zhu, Juntao Dai, Jiaming Ji, et al.2026 · 3 citations
11. InsightEval: An Expert-Curated Benchmark for Assessing Insight Discovery in LLM-Driven Data AgentsFindings of the Association for Computational LinguisticsZhenghao Zhu, Yuanfeng Song, Xing Chen, et al.2026 · 2 citations
12. How well can large language model agents simulate complex legal dispute resolution? ↗Artificial Intelligence and Law (Springer) · no summary yetYujin Zhou et al.2026 · 1 citation
13. ContextLens: Modeling Imperfect Privacy and Safety Context for Legal Compliance ↗Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · no summary yetHaoran Li et al.2026
14. LexAgentHallu: A Hierarchical Benchmark for Profiling Hallucinations in Legal Agents ↗arXiv:2609.09754 · EMNLP 2026 Main · no summary yetYujin Zhou et al.2026

🤝 Co-authors here

12 of 47 co-authors with a page, the most papers together first.

• Yike Guo ×6
• Jiaming Ji ×4
• Haoran Li ×3
• Boyuan Chen ×2
• Chengzhong Liu ×2
• Cui Yakun ×2
• Han Zhu ×2
• Hao Gu ×2
• Hao Wang ×2
• Huihao Jing ×2
• Juntao Dai ×2
• Peizhao Hu ×2

Venues

Where their works here were published, the most works first, with the editions of those that have a page.

• Annual Meeting of the Association for Computational Linguistics (ACL) ×9: ACL 2026 ×4, ACL 2025 ×4
• Conference on Empirical Methods in Natural Language Processing (EMNLP): EMNLP 2025
• Findings of the Association for Computational Linguistics: Findings of the Association for Computational Linguistics 2026
• Artificial Intelligence and Law (Springer)

🔗 Cited by

Papers of our library cite their works 125 times (of the 149,439 papers whose bibliographies we read); 29 citing papers with a page here →.

Updated 31 October 2026

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/authors/sirui-han/