A reference document used to keep track of the reading group meetings. Pull requests are welcome. Meeting Link
| Date | Presenter | Paper | Why should we care | Appear in | Repo | Blogs |
|---|---|---|---|---|---|---|
| 22/01/2025 | Jooyoung Lee | Are Large Language Models Capable of Generating Human-Level Narratives? | Analyses AI generated storylines through three discourse-level aspects: i) story arcs; ii) turning points; and iii) affective dimensions including arousal and valence | EMNLP'24 | ||
| 5/02/2025 | Piotr Brodka | Invited talk from our international visitor | ||||
| 19/02/2025 | Elaine Gong | DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning | DeepSeek’s first-generation reasoning models DeepSeek-R1-Zero and DeepSeek-R1 | arxiv | ||
| 5/03/2025 | Frankie Yuan | A primer on partially observable Markov decision processes (POMDPs) | A (somewhat) gentle introduction to Partially Observed Markov Decision Processes (POMDP) | Methods in Ecology and Evolution | ||
| 19/03/2025 | Emily Booth | Using thematic analysis in psychology | A basic introduction to one form of qualitative data analysis used widely by social science researchers | Qualitative Research in Psychology | ||
| 9/04/2025 | Frankie Yuan | Behavioral Homophily in Social Media via Inverse Reinforcement Learning: A Reddit Case Study | Frankie will dry-run his WWW presentation. | The Web Conference (WWW’25) | ||
| 16/04/2025 | Marian-Andrei Rizoiu | An overview of the thesis submission process, including Candidature assessments and selecting reviewers. | ||||
| 30/04/2025 | No reading group! Find us at WWW’25 | |||||
| 7/05/2025 | Philipp Schneider | Reinforcement-Learning Based Covert Social Influence Operations | Presents a reinforcement learning framework that models Covert Social Influence Operations on social media to optimise the actions of malicious agents in manipulating public opinion. | WWW'25 | ||
| 14/05/2025 | Philipp Schneider | The AI Economist: Taxation policy design via two-level deep multiagent reinforcement learning | Introduces a two-level deep reinforcement learning framework where economic agents and a social planner co-adapt to design optimal taxation policies. | Science | ||
| 21/05/2025 | Jooyoung Lee | Exposing Cross-Platform Coordinated Inauthentic Activity in the Run-Up to the 2024 U.S. Election | Introduces a coordination detection model to analyse cross-platform coordinated inauthentic activity across X (formerly Twitter) Facebook and Telegram during the 2024 U.S. Election. | WWW'25 | ||
| 4/06/2025 | Rohit Ram | Practical Guidelines for Ideology Detection Pipelines and Psychosocial Applications | Rohit will dry-run his ICWSM presentation | ICWSM'25 | ||
| 11/06/2025 | Cancelled | |||||
| 25/06/2025 | Huixiang Fu | The susceptibility paradox in online social influence | The paper investigates susceptibility to influence within social networks focusing on the differential effects of influence-driven versus spontaneous behaviors on user content adoption | ICWSM'25 | ||
| 9/07/2025 | Elaine Gong | Harnessing the Universal Geometry of Embeddings | An unsupervised approach translates any embedding to and from a universal latent representation | arXiv | ||
| 23/07/2025 | Lin Tian | Safety Alignment Should Be Made More Than Just a Few Tokens Deep | Outstanding Paper Award at ICLR2025. This paper critically examines why current Large Language Models (LLMs) remain vulnerable despite extensive safety alignment training. | ICLR'25 | ||
| 6/08/2025 | Cancelled due to technical issus | |||||
| 20/08/2025 | Frankie Yuan | A Closer Look at Machine Unlearning for Large Language Models | Examines the training and evaluation of machine unlearning techniques on LLMs. | ICLR'25 | ||
| 3/09/2025 | Lin Tian | Dry run of the CLEF’25 presentation on the Mario at EXIST paper. | The paper proposes a new pipeline that won the EXIST’25 challenge to identify hate speech. | CLEF’25 | ||
| 10/09/2025 | Cancelled | |||||
| 17/09/2025 | Emily Booth | The UTS Ethics process at UTS and beyond | The Ethics process is integral to research design; it makes sure we are following the “cans” and “cannots” of research, and acts as an insurance policy post publication. | |||
| 24/09/2025 | Marian-Andrei Rizoiu | Paper that Andrei reviewed for IEEE Transactions on Computational Social Systems | The review process; the session covered research from the other side: the reviewer. And lessons learned for our own submissions. | IEEE Transactions on Computational Social Systems | ||
| 1/10/2025 | Lin Tian | Debrief on CLEF’25 and German research visit | Lin presented at CLEF’25 and also did a one-week research visit in Germany, sponsored by DAAD. | |||
| 8/10/2025 | Cancelled | |||||
| 15/10/2025 | Jooyoung Lee | Summary of a Haystack: A Challenge to Long-Context LLMs and RAG Systems | Leverages the task of summarization as a testbed for evaluating long-context models and RAG systems. | EMNLP'24 | ||
| 22/10/2025 | Huixiang Fu | On Fake News Detection with LLM Enhanced Semantics Mining | The paper proposes a model based on LLM to depict the relations among news entities and topics as a heterogeneous graph for fake news detection task. | EMNLP'24 | ||
| 29/10/2025 | Andrew Savchenko | Empathy in Child-Robot Interaction: The Role of Narrative Framing, Age, Gender, and Baseline Empathy. | Group relevance, commendable level of transparency and academic rigour. | |||
| 5/11/2025 | Kevin Zheng | Two-Stream Convolutional Networks for Action Recognition in Videos | Introduces a dual-stream CNN architecture that fuses spatial and temporal cues to advance video action recognition. | NeurIPS'14 | ||
| 12/11/2025 | Md Zakir Hossain | Invited lecture by our research visitor. | ||||
| 19/11/2025 | Frankie Yuan | Consistently Simulating Human Personas with Multi-Turn Reinforcement Learning | This paper examines the reliability and consistency of LLM simulated human personas. | NeurIPS'25 | ||
| 26/11/2025 | Jooyoung Lee | |||||
| 3/12/2025 | Emily Booth | |||||
| 10/12/2025 | Majid Akhtar |