Long-term Measurements: Towards a Longitudinal Understanding of Human-AI Interactions
I will likely fail to keep this list regularly updated. Please see my
Google Scholar
for a more updated list.
Publications
Detecting and Controlling Sycophancy with Cascading Linear Features
Tailored to you: longitudinal effects of personalising language models
Positive Alignment: Artificial Intelligence for Human Flourishing
Is This Just Fantasy? Language Model Representations Reflect Human Judgments of Event Plausibility
Nuanced Safety for Generative AI: How Demographics Shape Responsiveness to Severity
Evaluating safety in GenAI Image safety annotation tasks from diverse annotators' perspectives
Utility-inspired Reward Transformations Improve Reinforcement Learning Training of Language Models
Nuanced Safety for Generative AI: How Demographics Shape Responsiveness to Severity
Whose view of safety? A deep DIVE dataset for pluralistic alignment of text-to-image models
Value Profiles for Encoding Human Variation
States as Strings as Strategies: Steering Language Models with Game-Theoretic Solvers
Skill Generalization with Verbs.
Pragmatics in grounded language learning: Phenomena, tasks, and modeling approaches.
Mapping Language Models to Grounded Conceptual Spaces.
Generalising to New Domains by Mapping Natural Language to Lifted LTL.
Does linguistic bias affect generative language models?
Game-theoretic Vocabulary Selection for Text Classification Tasks
Affordance-based Robot Object Retrieval
Room-Across-Room: Multilingual Vision-and Language Navigation with Dense Spatiotempral Grounding
On the Relationship Between Structure in Natural Language and Models of Sequential Decision Processes
Grounding Language to Non-Markovian Tasks with No Supervision of Task Specifications.
Robot Object Retrieval with Contextual Natural Language Queries.
How to Get Past Sesame Street: Sentence-Level Pretraining Beyond Language Modeling
Planning with State Abstractions for Non-Markovian Task Specifications
Learning Visually Grounded Meaning Representations with Sketches
Learning to Ground Language to Temporal Logical Form.
Probing What Different NLP Tasks Teach Machines about Function Word Comprehension
Looking for ELMo's Friends: Sentence-Level Pretraining Beyond Language Modeling.
Modeling Ambiguity in Text: A Corpus of Legal Literature.
A Corpus with Multi-Level Annotations of Patients, Interventions and Outcomes to Support Language Processing for Medical Literature.
Syntactic Patterns Improve Information Extraction for Medical Literature
Tutorials
Recognising Multimodal Entailment.
Lectures and Invited Talks
Columbia University, Data Science Institue
Princeton University, PACT Summer Program