Sarah Chen

Dr. Sarah Chen

Associate Professor of Computer Science

Stanford University

My research focuses on natural language processing, machine learning, and the intersection of language understanding with knowledge representation.

📍 Stanford, CA

🏛 Stanford University

✉ Email 🎓 Google Scholar 🆔 ORCID 💻 GitHub

Publications

You can also find my articles on my Google Scholar profile.

Journal Articles

Faithful Chain-of-Thought Reasoning via Semantic Entailment Verification

Published in Transactions of the ACL, 2026

We propose a verification framework that checks each step of a language model's chain-of-thought reasoning against a semantic entailment graph, improving faithfulness by 34% on multi-hop QA benchmarks while maintaining fluency.

Recommended citation: S. Chen, J. Park, M. Rodriguez, A. Gupta. (2026). "Faithful Chain-of-Thought Reasoning via Semantic Entailment Verification." Transactions of the ACL, 14, 312-329.

Download Paper | Download Slides

Robust Factuality Evaluation for Open-Domain Text Generation

Published in Journal of Machine Learning Research, 2025

We introduce FactScore-R, a robust evaluation metric for factual consistency in open-domain text generation that accounts for partial entailment and varying levels of granularity, achieving higher correlation with human judgments than existing metrics.

Recommended citation: A. Gupta, S. Chen. (2025). "Robust Factuality Evaluation for Open-Domain Text Generation." JMLR, 26(1), 1-45.

Download Paper | BibTeX

Scaling Knowledge Graph Completion with Contrastive Pre-Training

Published in Artificial Intelligence, 2024

We present a contrastive pre-training approach for knowledge graph completion that scales to graphs with millions of entities, achieving state-of-the-art results on FB15k-237 and WN18RR while requiring 4x less training compute.

Recommended citation: L. Wei, R. Nakamura, S. Chen. (2024). "Scaling Knowledge Graph Completion with Contrastive Pre-Training." Artificial Intelligence, 335, 104018.

Download Paper | BibTeX

Conference Papers

KnowledgeBridge: Integrating Structured Knowledge Graphs with Large Language Models

Published in ICLR, 2026

We introduce KnowledgeBridge, a framework for integrating structured knowledge graphs with large language models for multi-hop reasoning. Our approach dynamically retrieves and grounds KG subgraphs during generation, improving multi-hop QA accuracy by 18%.

Recommended citation: R. Nakamura, S. Chen, L. Wei. (2026). "KnowledgeBridge: Integrating Structured Knowledge Graphs with Large Language Models." In Proceedings of ICLR 2026.

Download Paper | Code

Cross-Lingual Semantic Parsing with Minimal Supervision

Published in EMNLP, 2025 — Oral Presentation

We present a cross-lingual transfer method for semantic parsing that requires only 50 annotated examples in the target language, leveraging multilingual pre-training and a novel alignment objective to bridge the gap between high- and low-resource languages.

Recommended citation: S. Chen, K. Okafor, T. Müller. (2025). "Cross-Lingual Semantic Parsing with Minimal Supervision." In Proceedings of EMNLP 2025.

Download Paper | Code | Download Slides

Neurosymbolic Approaches to Compositional Generalization in Semantic Parsing

Published in NeurIPS, 2024

We combine neural sequence-to-sequence models with symbolic grammar constraints to achieve compositional generalization in semantic parsing, outperforming pure neural approaches on COGS and SCAN benchmarks by a significant margin.

Recommended citation: J. Park, S. Chen. (2024). "Neurosymbolic Approaches to Compositional Generalization in Semantic Parsing." In Advances in NeurIPS 2024.

Download Paper | Code | Poster