About Me

I am Ting-Rui (Ray) Chiang, a PhD student at University of Southern California, working with Prof. Dani Yogatama. Previously, I stayed a while at Carnegie Mellon University (CMU) and National Taiwan University (NTU) advised by Prof. Yun-Nung (Vivian) Chen. My research interest is about language models, e.g., the mechanism behind the self-supervised training process.

GitHub   Google Scholar   Semantic Scholar  

Publications

The Distributional Hypothesis Does Not Fully Explain the Benefits of Masked Language Model Pretraining

Ting-Rui Chiang, Dani Yogatama

Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023, full paper)

[Slides] [Poster] [Source Code]

On a Benefit of Mask Language Modeling: The Robustness to Simplicity Bias

Ting-Rui Chiang

Proceedings of the 3nd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 13th International Joint Conference on Natural Language Processing (IJCNLP-AACL 2023, full paper, oral)

[Slides]

Breaking Down Multilingual Machine Translation

Ting-Rui Chiang, Yi-Pei Chen, Yi-Ting Yeh, Graham Neubig

Findings of the Association for Computational Linguistics: ACL 2022 (ACL 2022 Findings, full paper)

[Slides]

Relating Neural Text Degeneration to Exposure Bias

Ting-Rui Chiang, Yun-Nung Chen

Proceedings of the Fourth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP. (BlackboxNLP@EMNLP 2021)

[Poster] [Source Code]

Improving Dialogue State Tracking by Joint Slot Modeling

Ting-Rui Chiang, Yi-Ting Yeh

Proceedings of the 3rd Workshop on Natural Language Processing for Conversational AI. (NLP for ConvAI@EMNLP 2021, oral)

[Slides] [Source Code]

Are you doing what I say? On modalities alignment in ALFRED

Ting-Rui Chiang*, Yi-Ting Yeh*, Ta-Chung Chi and Yao-Shian Wang

Proceedings of the 1st Workshop on Novel Ideas in Learning-to-Learn through Interaction. (NILLI@EMNLP 2021, oral)

[Slides]

An Empirical Study of Content Understanding in Conversational Question Answering

Ting-Rui Chiang, Hao-Tong Ye and Yun-Nung Chen

Proceedings of Thirty-Fourth AAAI Conference on Artificial Intelligence, New York, New York, USA, 2020. (AAAI 2020, full paper, oral)

[Poster] [Source Code]

Semantically-Aligned Equation Generation for Solving and Reasoning Math Word Problems

Ting-Rui Chiang, Yun-Nung Chen

Proceedings of The 17th Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Minneapolis, USA, 2019. (NAACL-HLT 2019, full paper, oral)

[Slides] [Source Code]

Learning Multi-Level Information for Dialogue Response Selection by Highway Recurrent Transformer

Ting-Rui Chiang, Chao-Wei Huang, Shang-Yu Su, Yun-Nung Chen

Computer Speech & Language, 2020. (Journal)

[Source Code]

RAP-Net: Recurrent Attention Pooling Networks for Dialogue Response Selection

Ting-Rui Chiang*, Chao-Wei Huang*, Shang-Yu Su, Yun-Nung Chen

Computer Speech & Language, 2020. (Journal)

[Source Code]

Thoughts on Research

Though it was not about research, I found a paragraph by Eileen Chang describing my attitude toward research very well:

“This thing we call reality is without structure, a confusion of gramophones playing in chaotic cacophony, each singing its own song. But amid the unintelligible clamour is the unexpected lucid interval that sours the heart and moistens the eye, a discernible melody instantly reclaimed by the weighty gloom, the spark of understanding swamped. Painters, writers, and composers intertwine fragmentary, accidentally discovered harmonies, and attain artistic wholeness.” (From the Ash, translated by Oliver Stunt)

I hope I am going toward the wholeness.

Extracurricular Activities