Tim Althoff

Associate Professor, PI
Personal websitePublications
2026 · Nature Machine Intelligence
Capable language models can outgrow the benefits of collaboration
2026 · Proceedings of the International AAAI Conference on Web and Social Media
How Conversational Structure and Style Shape Online Community Experiences
2026 · Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Inferring Events from Time Series using Language Models
2026 · ICLR 2026 Workshop on Recursive Self-Improvement
Self-Improving VLM Judges Without Human Annotations
2026 · Nature Communications
Transforming wearable data into personal health insights using large language model agents
2025 · Proceedings of the ACM on Human-Computer Interaction
Perceptions of Moderators as a Large-Scale Measure of Online Community Governance
Best Paper Honorable Mention
2025 · Proceedings of the International AAAI Conference on Web and Social Media
Reddit Rules and Rulers: Quantifying the Link Between Rules and Perceptions of Governance Across Thousands of Communities
Best Paper Award
2024 · arXiv preprint
A Computational Framework for Behavioral Assessment of LLM Therapists
2024 · Findings of EMNLP
BLADE: Benchmarking Language Model Agents for Data-Driven Science
2024 · arXiv preprint
Correcting misinformation on social media with a large language model
2024 · Findings of EMNLP
Language Models Still Struggle to Zero-shot Reason about Time Series
2024 · Nature Mental Health
Rethinking technology innovation for mental health: framework for multi-sectoral collaboration
2023 · ACL
Cognitive Reframing of Negative Thoughts through Human-Language Model Interaction
Outstanding Paper Award
2023 · ACM IMWUT
GLOBEM: Cross-Dataset Generalization of Longitudinal Human Behavior Modeling
Distinguished Paper Award
2023 · Conference on Health, Inference, and Learning
Homekit2020: A Benchmark for Time Series Classification on a Large Mobile Sensing Dataset with Laboratory Tested Ground Truth of Influenza Infections
2023 · Nature Machine Intelligence
Human–AI collaboration enables more empathic conversations in text-based peer-to-peer mental health support
2023 · JMIR Formative Research
Mapping User Engagement in Digital Psychotherapy: An Integrative Engagement Model
2023 · Conference on Health, Inference, and Learning
Self-supervised Pretraining and Transfer Learning Enable Flu and COVID-19 Predictions in Small Mobile Sensing Datasets
2022 · Nature Communications
Disparate impacts on online information access during the Covid-19 pandemic
2022 · Nature Communications
Large-scale diet tracking data reveal disparate associations between food environment and diet
2022 · Annual Review of Public Health
Leveraging Mobile Technology for Public Health Promotion: A Multidisciplinary Perspective
2021 · ICWSM
Political Bias and Factualness in News Sharing across more than 100,000 Online Communities
Outstanding Paper (Analysis)
2021 · ICWSM
The Effect of Moderation on Online Mental Health Conversations
Outstanding Paper (Study Design)
2018 · NPJ Digital Medicine
Psychomotor function measured via online activity predicts motor vehicle fatality risk
News
Tim, Inna, Ashish, and Galen win 4 Best Paper awards at WWW, ICWSM, and NLP4IF!
Awards
Best Paper Honorable Mention
ACM CSCW
- Galen Weld
- Tim Althoff
- Perceptions of Moderators as a Large-Scale Measure of Online Community Governance
UW Digital Accessibility Team Award — CREATE Accessible Data Science and STEM Lecture Team
University of Washington
CREATE Accessible Data Science and STEM Lecture Team, including Tim Althoff
SourceDistinguished Paper Award
ACM UbiComp / IMWUT
- Yasaman Sefidgar
- Tim Althoff
- GLOBEM: Cross-Dataset Generalization of Longitudinal Human Behavior Modeling
Outstanding Paper (Analysis)
AAAI ICWSM
- Galen Weld
- Tim Althoff
- Political Bias and Factualness in News Sharing across more than 100,000 Online Communities
Best Paper Award
The Web Conference (WWW)
- Ashish Sharma
- Inna Lin
- Tim Althoff
- Towards Facilitating Empathic Conversations in Online Mental Health Support: A Reinforcement Learning Approach