🧑‍🔬 Lei Ji

🏛️ Affiliations

🏢Microsoft Research AsiaBeijing · China2020–2026 · 9 paperscurrent🏢MicrosoftRedmond · United States2020–2026 · 3 paperscurrent🏛️Chinese Academy of SciencesBeijing · China2010–2021 · 5 papers🎓University of Chinese Academy of SciencesBeijing · China2020–2021 · 5 papers🎓Beijing University of Posts and TelecommunicationsBeijing · China2017–2020 · 5 papers🏢Huawei TechnologiesShenzhen · China2023–2025 · 4 papers🔬Institute of Computing TechnologyBeijing · China2020–2021 · 4 papers🎓Southwest Jiaotong UniversityChengdu · China2021 · 3 papers

From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 2 more institutions 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. TL;DR: Too Long, Do Re-weighting for Efficient LLM Reasoning Compression (2025 · 8 citations)
   Reasoning-focused LLMs that generate long Chain-of-Thought (CoT) often overthink, spending many tokens even on easy questions. Existing compression approaches usually need laborious data labeling, model interpolation (as in CoT-Valve), or costly RL with length penalties.
2. Data Mixing Agent: Learning to Re-weight Domains for Continual Pre-training (2026 · 1 citation)
   Continual pre-training of an LLM on a narrow target field, such as math, often causes catastrophic forgetting of general skills.
3. DynamixSFT: Dynamic Mixture Optimization of Instruction Tuning Collections (2026)
   Instruction-tuning collections such as TÜLU combine many heterogeneous datasets, yet their mixing ratios are usually fixed by hand or need extra proxy models and validation sets to adjust.

All works

4 works of theirs in our library, the most cited first; 3 with a page here, the others link their original until we summarise them.

1. A Benchmark for Structured Procedural Knowledge Extraction from Cooking Videos ↗Proceedings of the First International Workshop on Natural Language Processing Beyond Text · no summary yetFrank F. Xu et al.2020 · 27 citations
2. TL;DR: Too Long, Do Re-weighting for Efficient LLM Reasoning CompressionarXiv:2506.02678Zhong-Zhi Li, Xiao Liang, Zihao Tang, et al.2025 · 8 citations
3. Data Mixing Agent: Learning to Re-weight Domains for Continual Pre-trainingAnnual Meeting of the Association for Computational Linguistics (ACL)Kailai Yang, Xiao Liu, Lei Ji, et al.2026 · 1 citation
4. DynamixSFT: Dynamic Mixture Optimization of Instruction Tuning CollectionsFindings of the Association for Computational LinguisticsHaebin Shin, Lei Ji, Xiao Liu, et al.2026

🤝 Co-authors here

12 of 17 co-authors with a page, the most papers together first.

• Xiao Liu ×3
• Yeyun Gong ×3
• Xiao Liang ×2
• Haizhen Huang
• Hao Li
• Haotian Xu
• Kailai Yang
• Mao Yang
• Peijie Wang
• Peng Cheng
• Qi Chen
• Weiwei Deng

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): ACL 2026
• Findings of the Association for Computational Linguistics: Findings of the Association for Computational Linguistics 2026
• Proceedings of the First International Workshop on Natural Language Processing Beyond Text

🔗 Cited by

Papers of our library cite their works 16 times (of the 149,439 papers whose bibliographies we read); 4 citing papers with a page here →.

Updated 1 November 2026

↑ back to top

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