Xin Chen
Xin Chen has 6 papers in our library, in 2014–2026, most of them on Language models & pretraining, Dialogue systems, Evaluation & benchmarks; often with Jingang Wang, Du Su, Fei Sun.
OpenAlex · h-index 0 · 6 works · 0 citations OpenAlex, 2026-10-14
📄 6 works in our library · 🔗 Cited 302 times in our library · ⭐ Paper Stars ranking: #3906
Affiliations
Fields, topics and concepts
📚 Shelves 12The field of science this work belongs to (OpenAlex fields).Arts and Humanities · 5 worksComputer Science · 5 worksArtificial Intelligence · 5 worksLanguage and Linguistics · 5 worksComputer Vision and Pattern Recognition · 1 workAtmospheric Science · 1 workBuilding and Construction · 1 workEarth and Planetary Sciences · 1 workEngineering · 1 workEnvironmental Engineering · 1 workEnvironmental Science · 1 workInformation Systems · 2 works
🧭 Topics 5Research subjects and languages the work is about.English · 1 workLanguage models & pretraining · 1 workDialogue systems · 1 workEvaluation & benchmarks · 1 workInformation retrieval & search · 1 work
Works by year
2020s
2010s
Papers
- InternLM2 Technical Report (2024 · 301 citations)
Open-source LLMs still trail proprietary systems such as GPT-4, and earlier technical reports rarely explain how pre-training data was prepared. - Pangu-Weather: A 3D High-Resolution Model for Fast and Accurate Global Weather Forecast (2022 · 133 citations)
Numerical weather prediction (NWP) solves physical equations on supercomputers. It is slow, which limits timeliness and ensemble size, and its parameterizations introduce errors. Earlier deep learning forecasters such as FourCastNet were fast but less accurate than ECMWF's operational IFS. - Automatic Assessment of the Speech of Young English Learners (2014 · 35 citations)
The paper describes how Pearson automated the speaking section of AZELLA, Arizona's yearly English proficiency test for K-12 English learners. - Too Consistent to Detect: A Study of Self-Consistent Errors in LLMs (2025 · 12 citations)
LLM error detectors often treat agreement across sampled answers as a sign of correctness. This paper identifies a blind spot in that assumption: self-consistent errors , where a model gives the same wrong answer, up to semantic equivalence, in its greedy output and in all stochastic samples. The authors define these… - BaseCal: Unsupervised Confidence Calibration via Base Model Signals (2026 · 1 citation)
Post-trained LLMs (PoLLMs), such as instruction-tuned or RLHF-aligned models, tend to be overconfident: they assign high probability even to wrong answers, which limits the usefulness of confidence for detecting hallucinations. - MTR-Suite: A Framework for Evaluating and Synthesizing Conversational Retrieval Benchmarks (2026)
Conversational retrieval decides which evidence a RAG system can use, but its benchmarks have two kinds of flaws.
Co-authors here
- Jingang Wang ×3
- Du Su ×2
- Fei Sun ×2
- Hexiang Tan ×2
- Xueqi Cheng ×2
- Xunliang Cai ×2
- Yuanzhuo Wang ×2
- Bei Li
- Haodong Duan
- Haojiong Chen
- Huawei Shen
- Jared Bernstein
Venues
- Annual Meeting of the Association for Computational Linguistics (ACL) ×2: ACL 2026 ×2
- Workshop on Innovative Use of NLP for Building Educational Applications (BEA): BEA 2014
- Conference on Empirical Methods in Natural Language Processing (EMNLP): EMNLP 2025
Cited by
Updated
This page in Markdown → https://md.research.elamur.ai/free/en/authors/xin-chen/