6 papers
Small Foundation Models of Human Cognition and Behaviour
Nick Oh, Fernand Gobet
Large language models fine-tuned on human behavioural data have emerged as general-purpose cognitive proxies, but the scale this requires, and whether these models process task str…
Reliability, Faithfulness, and the Limits of Post-hoc Explanations of Opaque Scientific Models
Nick Oh, Helen Jin
Post-hoc explanation methods are routinely used to interpret scientific machine learning models, with the deliverable understood to be insight into the phenomenon the model has bee…
Monitor-Generate-Verify (MGV): Formalising Metacognitive Theory for Language Model Reasoning
Nick Oh, Fernand Gobet
Test-time reasoning architectures such as those following the Generate-Verify paradigm, where a model iteratively refines or verifies its own generated outputs, prioritise generati…
In Defence of Post-hoc Explainability
Nick Oh
This position paper defends post-hoc explainability methods as legitimate tools for scientific knowledge production in machine learning. Addressing criticism of these methods' reli…
Before you <think>, monitor: Implementing Flavell's metacognitive framework in LLMs
Nick Oh
Current approaches to enhancing LLM reasoning follows two isolated paradigms: Monitor-Generate methods like Plan-and-Solve (Wang et al., 2023) and SELF-DISCOVER (Zhou et al., 2024)…
PETLP: A Privacy-by-Design Pipeline for Social Media Data in AI Research
Nick Oh, Giorgos D. Vrakas, Siân J. M. Brooke +2
Social media data presents AI researchers with overlapping obligations under the GDPR, copyright law, and platform terms -- yet existing frameworks fail to integrate these regulato…