I am a Research Engineer at Google DeepMind. I work on various machine learning problems that involve reasoning, planning, self-correction, and corresponding evaluation. Formerly, I have enjoyed internships at AI2 Aristo, Amazon A9, and Philips Research. I obtained my PhD Degree from the Kahlert School of Computing, University of Utah where I was very fortunate to be advised by Prof. Vivek Srikumar.
Selected Papers
-
Beyond Perplexity: Multi-dimensional Safety Evaluation of LLM Compression Permalink
Zhichao Xu, Ashim Gupta, Tao Li, Oliver Bentham, Vivek Srikumar ArXiv 2024
-
Revision Matters: Generative Design Guided by Revision Edits Permalink
Tao Li, Chin-Yi Cheng, Amber Xie, Gang Li, Yang Li ArXiv 2024
-
Devil’s Advocate: Anticipatory Reflection for LLM Agents Permalink
Haoyu Wang, Tao Li, Zhiwei Deng, Dan Roth, Yang Li. ArXiv 2024
-
Automatic Macro Mining from Interaction Traces at Scale Permalink
Forrest Huang, Gang Li, Tao Li, Yang Li CHI 2024 (oral presentation)
-
MUG: Interactive Multimodal Grounding on User Interfaces Permalink
Tao Li, Gang Li, Jingjie Zheng, Purple Wang, Yang Li. EACL Findings 2024
-
A Zero-Shot Language Agent for Computer Control with Structured Reflection Permalink
Tao Li, Gang Li, Zhiwei Deng, Bryan Wang, Yang Li. EMNLP Findings 2023 (oral presentation at NILLI)
-
Learning Semantic Role Labeling from Compatible Label Sequences Permalink
Tao Li, Ghazaleh Kazeminejad, Susan W. Brown, Martha Palmer, Vivek Srikumar. EMNLP Findings 2023
-
PYLON: A PyTorch Framework for Learning with Constraints Permalink
Kareem Ahmed, Tao Li, Thy Ton, Quan Guo, Kai-Wei Chang, Parisa Kordjamshidi, Vivek Srikumar, Guy Van den Broeck, Sameer Singh. NeurIPS Demo 2021 & AAAI 2022
-
OSCaR: Orthogonal Subspace Correction and Rectification of Biases in Word Embeddings Permalink
Sunipa Dev, Tao Li, Jeff M Phillips, Vivek Srikumar. EMNLP 2021
-
Automatic Entity State Annotation using the VerbNet Semantic Parser Permalink
Ghazaleh Kazeminejad, Martha Palmer, Tao Li, Vivek Srikumar. LAW-DMR at EMNLP 2021
-
UnQovering Stereotyping Biases via Underspecified Questions Permalink
Tao Li, Tushar Khot, Daniel Khashabi, Ashish Sabharwal and Vivek Srikumar. EMNLP Findings 2020
-
On Data Augmentation for Extreme Multi-label Classification Permalink
Danqing Zhang, Tao Li, Haiyang Zhang, Bing Yin. Arxiv 2020
-
Structured Tuning for Semantic Role Labeling Permalink
Tao Li, Parth Anand Jawale, Martha Palmer and Vivek Srikumar. ACL 2020
-
On Measuring and Mitigating Biased Inferences of Word Embeddings Permalink
Sunipa Dev, Tao Li, Jeff Philips and Vivek Srikumar. AAAI 2020 (oral presentation)
-
A Logic-Driven Framework for Consistency of Neural Models Permalink
Tao Li, Vivek Gupta, Maitrey Mehta and Vivek Srikumar. EMNLP 2019
-
Augmenting Neural Networks with First-order Logic Permalink
Tao Li and Vivek Srikumar. ACL 2019 (oral presentation)
-
Visual Interrogation of Attention-Based Models for Natural Language Inference and Machine Comprehension Permalink
Shusen Liu, Tao Li, Zhimin Li, Vivek Srikumar, Valerio Pascucci and Peer-Timo Bremer. EMNLP 2018 Demo
-
NLIZE: A Perturbation-Driven Visual Interrogation Tool for Analyzing and Interpreting Natural Language Inference Models Permalink
Shusen Liu, Zhimin Li, Tao Li, Vivek Srikumar, Valerio Pascucci and Peer-Timo Bremer. IEEE InfoVis 2018
-
Exploiting Sentence Similarities for Better Alignments Permalink
Tao Li and Vivek Srikumar. EMNLP 2016 (oral presentation)
Thesis
- Neural Learning with Rules for Data Efficiency
Tao Li. University of Utah. 2022
Experience
- May 2022 - Present
- Research Engineer, Google DeepMind
- Jan 2022 - May 2024
- Research Engineer, Google Research
- Jan 2020 - May 2020
- Research Intern, AllenAI Aristo
- May 2019 - Aug 2019
- Applied Scientist Intern, Amazon A9
- May 2018 - Aug.2018
- Research Intern, Philips Research
Talks
-
A Zero-Shot Language Agent for Computer Control with Structured Reflection
Talk at NILLI@EMNLP 2023, Singapore
-
End-to-end Neural Learning with Rules
Talk at ML Seminar, UCLA
-
Structured Tuning for Semantic Role Lableing
Talk at the Computational Semantics group, CU Boulder
-
Augmenting Neural Networks with First-order Logic
Talk at Session: Machine Learning, ACL 2019, Florence, Italy
-
Exploiting Sentence Similarities for Better Alignments
Talk at Session: Semantic Similarity, EMNLP 2016, Austin, Texas
Repositories
Check out my repos at:
Teaching
Services
- Area Chair for ACL Rolling Review.
- Routine reviewer for AAAI, AACL, ACL, ARR, CoNLL, EMNLP, ICLR, JMLR, NeurIPS, TKDE, and etc.