Papers by David Peng

FOLIO: Natural Language Reasoning with First-Order Logic (2024.emnlp-main)

Copied to clipboard

Simeng Han, Hailey Schoelkopf, Yilun Zhao, Zhenting Qi, Martin Riddell, Wenfei Zhou, James Coady, David Peng, Yujie Qiao, Luke Benson, Lucy Sun, Alexander Wardle-Solano, Hannah Szabó, Ekaterina Zubova, Matthew Burtell, Jonathan Fan, Yixin Liu, Brian Wong, Malcolm Sailor, Ansong Ni, Linyong Nan, Jungo Kasai, Tao Yu, Rui Zhang, Alexander Fabbri, Wojciech Kryscinski, Semih Yavuz, Ye Liu, Xi Lin, Shafiq Joty, Yingbo Zhou, Caiming Xiong, Rex Ying, Arman Cohan, Dragomir Radev

Reasoning

Challenge:	Existing benchmarks for logical reasoning in large language models lack language naturalness or limited complexity.
Approach:	They propose to use first-order logic annotations to evaluate logical reasoning capabilities of large language models.
Outcome:	The proposed dataset evaluates the FOL reasoning ability of supervised fine-tuning on medium-sized language models.

Unlocking Parameter-Efficient Fine-Tuning for Low-Resource Language Translation (2024.findings-naacl)

Copied to clipboard

Tong Su, Xin Peng, Sarubi Thillainathan, David Guzmán, Surangika Ranathunga, En-Shiun Lee

Responsible & Trustworthy NLP

Challenge:	Parameter-efficient fine-tuning (PEFT) methods are important in low-resource language (LRL) Neural Machine Translation (NMT) but their practical effectiveness varies significantly across different languages.
Approach:	They evaluated the performance of 8 parameters-efficient fine-tuning methods with 15 architectures using the SacreBLEU score.
Outcome:	The Houlsby+Inversion adapter outperforms the baseline architectures in both in-domain and out-domain tests and the Houlson+Inverter achieves the best performance overall.

Testing Paraphrase Models on Recognising Sentence Pairs at Different Degrees of Semantic Overlap (2023.starsem-1)

Copied to clipboard

Qiwei Peng, David Weir, Julie Weeds

Text Generation

Challenge:	Existing models for paraphrase detection are not suitable for many applications . existing datasets ignore and fail to test models in this setup .
Approach:	They propose to use adversarial paradigms to test paraphrase detection models . they propose to examine the sensitivity to different degrees of semantic overlap .
Outcome:	Empirical results show that paraphrase models and different sentence encoders appear successful on evaluations, but measuring the degree of semantic overlap remains a big challenge for them.

Predicate-Argument Based Bi-Encoder for Paraphrase Identification (2022.acl-long)

Copied to clipboard

Qiwei Peng, David Weir, Julie Weeds, Yekun Chai

Language Models

Challenge:	Cross-encoders have been widely applied to sentence pair tasks . bi-encodings have significantly lower computation complexity and are better suited to symmetric tasks.
Approach:	They propose a bi-encoder approach to the paraphrase identification task and incorporate predicate-argument information into SBERT.
Outcome:	The proposed model outperforms SBERT/SRoBERTa on six paraphrase identification datasets with a minimal increase in parameters.

VoiceCraft: Zero-Shot Speech Editing and Text-to-Speech in the Wild (2024.acl-long)

Copied to clipboard

Puyuan Peng, Po-Yao Huang, Shang-Wen Li, Abdelrahman Mohamed, David Harwath

Multimodality

Challenge:	VoiceCraft is a token-infilling neural codec language model for speech editing and zero-shot text-to-speech evaluation.
Approach:	They introduce a token infilling neural codec language model that performs on speech editing and zero-shot text-to-speech tasks.
Outcome:	The proposed model outperforms previous models on speech editing and zero-shot text-to-speech tasks.

Towards Structure-aware Paraphrase Identification with Phrase Alignment Using Sentence Encoders (2022.coling-1)

Copied to clipboard

Qiwei Peng, David Weir, Julie Weeds

Semantic Text Processing

Challenge:	Existing paraphrase identification datasets exhibit high correlation between positive pairs and the degree of their lexical overlap.
Approach:	They propose to combine sentence encoders with an alignment component by representing each sentence as a list of predicate-argument spans and decomposing the sentence-level meaning comparison into the alignment between their spans.
Outcome:	The proposed approach improves performance and interpretability for various sentence encoders.

Language Models Hallucinate, but May Excel at Fact Verification (2024.naacl-long)

Copied to clipboard

Jian Guan, Jesse Dodge, David Wadden, Minlie Huang, Hao Peng

Language Models

Challenge:	Recent advances in large language models (LLMs) have produced non-factual outputs . however, current LLMs suffer from the hallucination issue .
Approach:	They propose to use instruction-tuned LLMs to generate factual outputs . they find that FLAN-T5-11B performs best as a fact verifier .
Outcome:	The proposed method outperforms more capable LLMs like GPT3.5 and ChatGPT in the human evaluation.

Representing Syntax and Composition with Geometric Transformations (2021.findings-acl)

Copied to clipboard

Lorenzo Bertolini, Julie Weeds, David Weir, Qiwei Peng

Syntactic Text Processing

Challenge:	Existing models of word meaning are based on syntactic rather than proximal co-occurrences, but they are not suitable for syntax sensitive composition.
Approach:	They propose to encode syntactic structure by extending the Skip-Gram with Negative sampling architecture from word2vec.
Outcome:	The proposed models perform favourably on benchmark word similarity tasks on similarity tests on similar words compared to models based on proximal co-occurrence . however, the real promise of distributional models is the potential for syntax-sensitive composition.

VoiceStar: Robust Zero-Shot Autoregressive TTS with Duration Control and Extrapolation (2026.findings-acl)

Copied to clipboard

Puyuan Peng, Zhisheng Zheng, Shang-Wen Li, Abdelrahman Mohamed, David Harwath

Responsible & Trustworthy NLP

Challenge:	Neural codec language models (NCLMs) lack fine-grained controllability and inability to extrapolate to sequence lengths much longer than those seen during training.
Approach:	They propose a novel autoregressive encoder-decoder neural codec language model that can be trained with a Continuation-Prompt Mixed training system.
Outcome:	The proposed model outperforms or is on par with current state-of-the-art models on short-form benchmarks such as LibriSpeech and Seed-TTS in terms of intelligibility and naturalness.

Scaling Collaborative Effort with Agents (2026.findings-acl)

Copied to clipboard

Shannon Zejiang Shen, Valerie Chen, Ken Gu, Alexis Ross, Zixian Ma, Jillian Ross, Alex Gu, Chenglei Si, Wayne Chi, Andi Peng, Jocelyn J Shen, Ameet Talwalkar, Tongshuang Wu, David Sontag

Challenge:	Current evaluations of agents focus on producing high-quality, final outputs in one shot, failing to account for the inherently iterative nature of many real-world problems.
Approach:	They propose a framework that captures how an agent’s utility grows with increasing user involvement.
Outcome:	The proposed framework captures how an agent’s utility grows with increasing user involvement, revealing a missing ingredient in agent design: the ability to sustain engagement and scaffold user understanding.

VoiceCraft-X: Unifying Multilingual, Voice-Cloning Speech Synthesis and Speech Editing (2025.emnlp-main)

Copied to clipboard

Zhisheng Zheng, Puyuan Peng, Anuj Diwan, Cong Phuoc Huynh, Xiaohang Sun, Zhu Liu, Vimal Bhat, David Harwath

Multimodality

Challenge:	Autoregressive language model for multilingual speech editing and zero-shot text-to-speech synthesis is available in 11 languages.
Approach:	They introduce an autoregressive neural codec language model which unifies multilingual speech editing and zero-shot text-to-speech synthesis across 11 languages.
Outcome:	The model generates high-quality, natural-sounding speech, even with limited per-language data . it shows robust performance in diverse linguistic settings, even in limited per language data compared to other models .