基本信息
任昱冰 女 硕导 中国科学院信息工程研究所
电子邮件: renyubing@iie.ac.cn
通信地址: 北京市海淀区树村路19号中国科学院信息工程研究所
课题组主页:https://ascii-iie.github.io/
About me:https://lilice-r.github.io/
研究领域
研究方向为自然语言处理、大模型安全,目前聚焦于研究大模型水印和AI文本检测溯源。
招生信息
招收2027秋季学硕一位,欢迎联系,要求:
1. CET-6以上,具有vibe coding能力,擅长使用工具;
2. 内驱力和执行力强,做事有责任心,能够闭环完成每一件事;
3. 情绪稳定,能够与课题组的老师同学们一起合作,共同进步。
教育背景
2019-08--2024-06 中国科学院信息工程研究所 工学博士
2015-09--2019-07 北京化工大学 工学学士
工作经历
2024-07~现在, 中国科学院信息工程研究所, 预聘副高
教授课程
自然语言处理实战
CCF-A类论文
- Towards Reliable Marking and Verification of AI-Generated Text via Geometry-aware Sentence-level Watermarking. ICML'26. 第1作者.
- Rethinking LLM Watermark Detection in Black-Box Settings: A Non-Intrusive Third-Party Framework. Findings of ACL'26. 通讯作者.
- DualGuard: Dual-stream Large Language Model Watermarking Defense against Paraphrase and Spoofing Attack. Findings of ACL'26. 第2作者.
- Exons-Detect: Identifying and Amplifying Exonic Tokens via Hidden-State Discrepancy for Robust AI-Generated Text Detection. ACL'26. 通讯作者.
- Cognitive Analysis Graph-Guided Multi-Turn Safety Enhancement for Large Language Models. Findings of ACL'26. 第4作者.
- DNA-DetectLLM: Unveiling AI-Generated Text via a DNA-Inspired Mutation-Repair Paradigm. Spotlight of NeurIPS'25. 通讯作者.
- Exploring Polyglot Harmony: On Multilingual Data Allocation for Large Language Models Pretraining. NeurIPS'25. 第2作者
- PIG: Privacy Jailbreak Attack on LLMs via Gradient-based Iterative In-Context Optimization. ACL'25. 通讯作者.
- Reliably Bounding False Positives: A Zero-Shot Machine-Generated Text Detection Framework via Multiscaled Conformal Prediction. ACL'25. 通讯作者.
- From Trade-off to Synergy: A Versatile Symbiotic Watermarking Framework for Large Language Models. ACL'25. 通讯作者.
- Bridging the Gap: Aligning Language Model Generation with Structured Information Extraction via Controllable State Transition. Oral at WWW'25. 第2作者.
- Subtle Signatures, Strong Shields: Advancing Robust and Imperceptible Watermarking in Large Language Models. Findings of ACL'24. 第1作者.
- Query in Your Tongue: Reinforce Large Language Models with Retrievers for Cross-lingual Search Generative Experience. WWW'24. 第4作者
- Steering Large Language Models for Cross-lingual Information Retrieval. SIGIR'24. 第2作者.
- Retrieve-and-Sample: Document-level Event Argument Extraction via Hybrid Retrieval Augmentation. ACL'23. 第1作者.