Research Team

Students, Mentees & Collaborators

The students I supervise and mentor across artificial intelligence, cognitive science, and linguistics, the theses and papers we've produced together, and the groups and collaborators I work with.

Supervision & Mentorship

Students & Mentees

Mentoring is the part of the job I take most seriously — and it produces real research. Several supervised projects have become papers at EMNLP, BlackboxNLP and NeurIPS workshops.

MPhil Thesis Supervision · Advanced Computer Science (ACS) & MLMI

Bianca-Mihaela GanescuMPhil ACS · 2024–25

Thesis: Integrating Cognitively-Inspired Selective Attention Cues in Small Vision-Language Models. Co-supervised with Dr Andrew Caines and Prof Paula Buttery.

Looking to Learn: Token-wise Dynamic Gating for Low-Resource Vision-Language Modelling — BabyLM @ EMNLP 2025 · Outstanding Paper Award Thesis report (PDF)

Yeji HeoMPhil ACS · 2025–26

Cultural Knowledge in Pretrained Language Models through Behaviour, Neuron Localisation, and Pretraining Regimes — BlackboxNLP 2026 (under review)

Adam El-KholyMPhil MLMI · 2025–26

From Entropy to Exploration: What Limits Post-Training under GRPO? — NeurIPS 2026 Workshop (Pre-to-Post)

Undergraduate Research Students & UROP

Ellie Polyakova ReedUndergraduate · 2025, 2026

UROP: Evaluating Typological Effects of L2 English in Language Model Benchmarks (2025).

Claude Reinvestigates Herself: syntactic-agreement units in LLMs (replication) — BlackboxNLP 2026 Shared Units, Divergent Trajectories: Syntactic Representations in Bilingual & Multilingual LMs — MRL @ EMNLP 2026 UROP report (PDF)

Shivan AroraUndergraduate · 2025, 2026

UROP: Controlled Text Generation: a teacher–student paradigm for CEFR-levelled complexity control (2025).

Code-Switching as a Probe of Cross-Lingual Knowledge Transfer and Cultural Localisation — MRL @ EMNLP 2026 Claude Reinvestigates Herself (co-author) — BlackboxNLP 2026

Mingchuan ZouUndergraduate · 2026

Can We Trust Small Language Models as Judges? — NeurIPS 2026 Workshops (TAE / JUDGe) Development of Linear Truth Encodings in Language Models (replication) — BlackboxNLP 2026 Execution-Aware Routing for On-Device SLM Agents — NeurIPS 2026 (SLM-Agents)

Jacy ToUndergraduate

Abdullah Akram · Alisa PotomkivaUndergraduate · 2026

Visiting Students, Interns & Research Assistants

Andrzej SzablewskiVisiting student

LangMAP: A Language-Adaptive Approach to Tokenization — EMNLP 2026 (co-author)

Ali KheirkhahPicoLM Research Intern · Summer 2025

Co-supervised with Richard Diehl Martinez; mentored through the Google DeepMind Research Ready programme.

ALTA Institute Research Assistants2025–26

Line-managed together with Prof Andrew Caines and Prof Paula Buttery.

Project Supervision

Proposals, Theses & Supervised Reports

Project proposals for prospective MPhil students, together with materials from theses, UROP projects and research reports I have supervised or co-authored.

MPhil Project Proposals 2025–26

Proposed MPhil (Advanced Computer Science / MLMI) project topics in small language models, multilinguality and interpretability.

Project Proposals 2025–26

Supervised Projects & Reports

Selected reports and write-ups from MPhil theses, UROP projects and research projects I have supervised or co-authored.

MPhil Project (Ganescu) UROP Report (Reed) Argument Mining Multimodal Grounding Project

Collaborators

Groups & Collaborators

I am very grateful to a number of academics, researchers and PhD students whom I am currently working with and/or previously worked with.

The Cambridge NLIP / ALTA team — The Cambridge team I work alongside.

The Cambridge PicoLM Team

Richard Diehl Martinez (and his MPhil/Part III students: Yuval Weiss and David Demitri Africa), Ryan Daniels (a Machine Learning Engineer with the Accelerate Programme for Scientific Discovery), and Prof Paula Buttery.

Cambridge NLIP Group

Zeb Goriely, Pietro Lesci, Julius Cheng, Dr Guy Emerson, Dr Fermin Moscoso del Prado Martin – we are working on information-theoretic models of diachronic phonological typology. Paul Siewert and I have also been thinking about Category Theoretic approaches in Linguistics. Mila Marcheva, Xiaochen (Neo) Zhu.

ALTA Institute (Computer Science & Technology)

Gabrielle Gaudeau, Dr Diana Galvan Sosa, Dr Zheng Yuan, Yuan Gao.

XFACT (KAIST AI) & NAVER Cloud

Jiwoo Hong (KAIST/Amazon Rufus), Noah Lee (KAIST/Naver Cloud), Dr James Thorne, Jeonghoon Kim (NAVER Cloud, KAIST AI), Woojin Chung.

Cambridge Theoretical & Applied Linguistics (TAL)

Nuria Bosch Masip.

Cambridge Language Technology Lab (LTL)

Yijie Zhou (EJ), Fangyu Liu, Prof Nigel Collier.

Individual Collaborators

Dr Donya Rooin (MilaNLP, Bocconi University) and Hongyi (Sam) Gu (KCL/NetMind.AI). Dr Konstantinos Voudouris (Institute for Human-Centered AI at Helmholtz Munich). Marek Masiak (Oxford University). Dr Thiemo Wambsganss (University of St Gallen, now Bern University – UROP Supervisor 2021).

Team dinner — A team dinner on the road.

Acknowledgements

Papers I have been acknowledged in:

Goriely & Buttery (2025), IPA-CHILDES & G2P+.