papers

Publications (14)

cs.CL2026

LongBench Pro: A More Realistic and Comprehensive Bilingual Long-Context Evaluation Benchmark

Ziyang Chen, Xing Wu, Junlong Jia +4

The rapid expansion of context length in large language models (LLMs) has outpaced existing evaluation benchmarks. Current long-context benchmarks often trade off scalability and r…

cs.CL2025

EntropyLong: Effective Long-Context Training via Predictive Uncertainty

Junlong Jia, Ziyang Chen, Xing Wu +5

Training long-context language models to capture long-range dependencies requires specialized data construction. Current approaches, such as generic text concatenation or heuristic…

cs.CL2025

NExtLong: Toward Effective Long-Context Training without Long Documents

Chaochen Gao, Xing Wu, Zijia Lin +2

Large language models (LLMs) with extended context windows have made significant strides yet remain a challenge due to the scarcity of long documents. Existing methods tend to synt…

cs.LG2026

PolicyLong: Towards On-Policy Context Extension

Junlong Jia, Ziyang Chen, Xing Wu +4

Extending LLM context windows is hindered by scarce high-quality long-context data. Recent methods synthesize data with genuine long-range dependencies via information-theoretic ve…

cs.CL2022

InfoCSE: Information-aggregated Contrastive Learning of Sentence Embeddings

Xing Wu, Chaochen Gao, Zijia Lin +3

Contrastive learning has been extensively studied in sentence embedding learning, which assumes that the embeddings of different views of the same sentence are closer. The constrai…

cs.CL2022

Text Smoothing: Enhance Various Data Augmentation Methods on Text Classification Tasks

Xing Wu, Chaochen Gao, Meng Lin +3

Before entering the neural network, a token is generally converted to the corresponding one-hot representation, which is a discrete distribution of the vocabulary. Smoothed represe…