Xiaojun Wan
研究员
Researches natural language processing and text mining, including text generation, summarization and LLM evaluation.
Xiaojun Wan's lab at Peking works on dialogue, generation & summarization and large language models. 74 papers since 2023, 44 at top venues, mostly ACL and EMNLP. 7 PhD students and 2 master's students. 44 alumni, most often to Alibaba and ByteDance. Applications close .
At a glance
- papers, 2023–2026
- 19/ year
- at T1 venues
- 59%
- PhD students
- 7
- MS students
- 2
- alumni listed
- 44
Papers by yearfrom the lab homepage, 2023–2026
- 21 papers2026 · T1 14 · collecting
- 22 papers2025 · T1 15
- 16 papers2024 · T1 7
- 15 papers2023 · T1 8
T1 venuesOtherVenues: ACL, EMNLP, AAAI, ICML, LREC
Research focuspaper titles that use each topic's words
- Large language models35 papers
- text10 papers
- Dialogue, generation & summarization9 papers
- summarization3 papers
- text generation1 paper
Paper list
From the lab homepage.
202621 papers
- ACL (Findings).CE-RM: A Pointwise Generative Reward Model Trained with Two-Stage Rollout and Unified Criteria
- ACL (T1)DecoCal: Decoding with Calibration in Diffusion Large Language Models
- ICML (T1)Dynamic Multimodal Evaluation via Knowledge-Enhanced Benchmark Evolution
- TOMMEAMA: Entity-Aware Multimodal Alignment Based Approach for News Image Captioning
- ACL (T1)Edit-Aware Reward Modeling for Chinese Grammatical Error Correction
- Findings of EMNLPFactNLI: Dynamic and Automated Fact-Based Augmentation of NLI Benchmarks
- ACL (T1)From TDMA to CDMA: A Multi-bit Watermark for Diffusion Language Models
- ACL (Findings).Ghost in the Shell: Synonym-Aware Logit Shaping Fingerprint for Copyright Protection of Large Vision-Language Models
- ACL (T1)HAD: HAllucination Detection Language Models Based on a Comprehensive Hallucination Taxonomy
- ACL (Findings).JointCQ: Improving Factual Hallucination Detection with Joint Claim and Query Generation
- ACL (T1)LEDOM: Reverse Language Model
- ACL (Findings).Minos: A Multimodal Evaluation Model for Bidirectional Generation Between Image and Text
202522 papers
- ACL (T1)A Dual-Perspective NLG Meta-Evaluation Framework with Automatic Benchmark and Better Interpretability
- NAACL (T1)Analyzing and Evaluating Correlation Measures in NLG Meta-Evaluation
- NAACL (T1)B4: A Black-Box Scrubbing Attack on LLM Watermarks
- CIKM (T1)C-FAITH: A Chinese Fine-Grained Benchmark for Automated Hallucination Evaluation
- EMNLP (T1)DAMON: A Dialogue-Aware MCTS Framework for Jailbreaking Large Language Models
- AAAI (T1)DSGram: Dynamic Weighting Sub-Metrics for Grammatical Error Correction in the Era of Large Language Models
- NeurIPS (T1)Enhancing LLM Watermark Resilience Against Both Scrubbing and Spoofing Attacks
- NAACL (Findings).Evaluating Self-Generated Documents for Enhancing Retrieval-Augmented Generation with Large Language Models
- EMNLP (Findings).Exploring and Evaluating Multimodal Knowledge Reasoning Consistency of Multimodal Large Language Models
- CIKM (T1)Exploring Causal Effect of Social Bias on Faithfulness Hallucinations in Large Language Models
- ACL (T1)Gödel Agent: A Self-Referential Agent Framework for Recursively Self-Improvement
- ACL (T1)ICR Probe: Tracking Hidden State Dynamics for Reliable Hallucination Detection in LLMs
202416 papers
- —Are LLM-based Evaluators Confusing NLG Quality Criteria? ACL 2024. (Long Paper)
- ACL (T1)Benchmarking Knowledge Boundary for Large Language Model: A Different Perspective on Model Evaluation
- AAAI (T1)Better than Random: Reliable NLG Human Evaluation with Constrained Active Sampling
- LRECContextual Modeling for Document-level ASR Error Correction
- NAACL (T1)Defining and Detecting Vulnerability in Human Evaluation Guidelines: A Preliminary Study Towards Reliable NLG Evaluation
- WACVDescribe Images in a Boring Way: Towards Cross-Modal Sarcasm Generation
- ACL (T1)Enhancing Large Language Models in Coding Through Multi-Perspective Self-Consistency
- LRECError-Robust Retrieval for Chinese Spelling Check
- AAAI (T1)History Matters: Temporal Knowledge Editing in Large Language Model
- LRECImage Matters: A New Dataset and Empirical Study for Multimodal Hyperbole Detection
- LRECIs Summary Useful or Not? An Extrinsic Human Evaluation of Text Summaries on Downstream Tasks
- EMNLP (Findings).PaCoST: Paired Confidence Significance Testing for Benchmark Contamination Detection in Large Language Models
202315 papers
- EMNLP (Findings).A New Benchmark and Reverse Validation Method for Passage-level Hallucination Detection
- ACL (T1)A New Dataset and Empirical Study for Sentence Simplification in Chinese
- EMNLP (T1)ALCUNA: Large Language Models Meet New Knowledge
- ACL (Findings).Evaluating Factuality in Cross-lingual Summarization
- EACLExploiting Summarization Data to Help Text Simplification
- EMNLP (Findings).Exploring Context-Aware Evaluation Metrics for Machine Translation
- EMNLP (T1)Exploring Discourse Structure in Document-level Machine Translation
- ACL (Findings).Exploring the Impact of Vision Features in News Image Captioning
- ACL (T1)MIL-Decoding: Detoxifying Language Models at Token-Level via Multiple Instance Learning
- EMNLP (T1)Models See Hallucinations: Evaluating the Factuality in Video Captioning
- EMNLP (Findings).New Datasets and Controllable Iterative Data Augmentation Method for Code-switching ASR Error Correction
- ACL (T1)Reference Matters: Benchmarking Factual Error Correction for Dialogue Summarization with Fine-grained Evaluation Framework
Where alumni went44 listed, 86% to industry
Industry 25Academia 4
Most often to
- Alibaba
- ByteDance
- Microsoft
Also to
- UCLA
Apply to PekingFall 2027
Computer Science PhD (international applicants)
- Round
- Main
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- CNPeking
Houfeng Wang
- Knowledge, QA & retrieval
- CNPeking
Zhifang Sui
- Knowledge, QA & retrieval
- USNYU
Eunsol Choi
- Knowledge, QA & retrieval
- GBEdinburgh
Emily Allaway
- Large language models
- USUT Austin
Junyi Jessy Li
- Dialogue, generation & summarization
- USNYU
Greg Durrett
- Large language models
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
Pekingcloses