Conference Programme

Time Zone: AEDT (UTC+11)

Wednesday November 26th - PNR Learning Studio 310

Time	Event
14:00	Tutorial
15:30	Afternoon Tea
16:00	Tutorial
17:00	Informal Meet-n-Greet

Thursday, November 27 - PNR Lecture Theatre 302

Time	Event
09:10	Opening & Welcome
09:30	Lightning Talks	Benchmarks & Evaluation
10:00	Keynote	Alistair Knott
11:00	Morning Tea
11:30	Talks	Robustness and Safety of Language Models
12:30	Lunch
14:00	Panel Discussion	Teaching NLP/Using NLP for Teaching
15:00	Afternoon Tea
15:30	Talks	Multimodality and Language
16:45	Lightning Talks	Hallucination & Vulnerabilities
17:30	Social Event	Friend in Hand Hotel, Top Bar, 58 Cowper St, Glebe, NSW, 2037

Friday, November 28 - PNR Lecture Theatre 302

Time	Event
09:00	Lightning Talks	Model Architecture & Learning
09:30	Talks	Language Generation
10:30	Morning Tea
11:00	Talks	Language Understanding
11:30	Shared Task	Normalise Adverse Drug Events
12:30	Lunch
14:00	Keynote	Mark Johnson
15:00	Afternoon Tea
15:30	ALTA General Meeting
16:00	Awards and Closing

Wednesday November 26th

Tutorial, 2pm - 5pm

Title: Alignment of Large Language Models with Human Preferences and Values

Location: PNR Learning Studio 310, location, and about the space

Livestream: The tutorial will be livestreamed here. Note, the rest of the workshop is not being livestreamed.

Presenters: Usman Naseem, Gautam Siddharth Kashyap, Kaixuan Ren, Yiran Zhang, Utsav Maskey, Juan (Ada) Ren, Afrozah Nadeem (Macquarie University)

Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities, yet their reliability and alignment with human expectations remain unresolved challenges. This tutorial introduces the foundations of alignment and provides participants with a conceptual and practical understanding of the field. Core principles such as values, safety, reasoning, and pluralism will be presented through intuitive explanations, worked examples, and case studies. The aim is to equip attendees with the ability to reason about alignment goals, understand how existing methods operate in practice, and critically evaluate their strengths and limitations.

Pre-requisite: A working knowledge of basic machine learning and NLP methods will help participants follow the technical material, but the tutorial is structured to remain accessible to those without prior alignment experience.

Content Outline:

Foundations of Alignment
Value and Preference Alignment for Helpfulness
Safety and Honesty Alignment
Pluralistic and Cross-Modal Alignment

Informal Meet-n-Greet, 5pm - 6:30pm

Right after the tutorial, there will be light food and drinks. Hang out, catch up, and meet new people!

Thursday, November 27

Location: PNR Lecture Theatre (2) 302

09:10: Opening & Welcome
09:20: Programme Committee Summary
09:30: Session #1 - Lightning Talks (Topic: Benchmarks & Evaluation, Session Chair: Qiongkai Xu)
- BESSTIE: A Benchmark for Sentiment and Sarcasm Classification for Varieties of English
  - Lead Author: Dipankar Srirag
- Evaluating Evidence Attribution in Generated Fact Checking Explanations
  - Lead Author: Rui Xing
- ExtremeAIGC: Benchmarking LMM Vulnerability to AI-Generated Extremist Content
  - Lead Author: Bhavik Chandna
- Good Intentions Beyond ACL: Who Does NLP for Social Good, and Where?
  - Lead Author: Grace LeFevre
  - Presenter: Denis Peskoff
- The More, The Better? A Critical Study of Multimodal Context in Radiology Report Summarization
  - Lead Author: Mong Yuan Sim
- Why Do Self-Harm Prediction Models Struggle to Generalise? – Lexical and Semantic Variations
  - Lead Author: Liuliu Chen
10:00: Keynote #1 - Alistair Knott
11:00: Morning Tea
11:30: Session #2 - Oral Presentations (Topic: Robustness and Safety of Language Models, Session Chair: Zhuang Li)
- Robustness of Neurosymbolic Reasoners on First-Order Logic Problems
  - Lead Author: Hannah Bansal
- A Dataset and Benchmark on Extraction of Novel Concepts on Trust in AI from Scientific Literature
  - Lead Author: Melanie McGrath
  - Presenter: Necva Bolucu
- SHIELD: Classifier-Guided Prompting for Robust and Safer LVLMs
  - Lead Author: Juan Ren
- Some Odd Adversarial Perturbations and the Notion of Adversarial Closeness
  - Lead Author: Shakila Mahjabin Tonni
12:30: Lunch
14:00: Panel Discussion - “Teaching NLP/Using NLP for Teaching” (Speakers: Massimo Piccardi, Jey Han Lau, Thomas Elton, Moderator: Aditya Joshi)
15:00: Afternoon Tea
15:30: Session #3 - Oral Presentations (Topic: Multimodality and Language, Session Chair: Jonathan K. Kummerfeld)
- Graph-Score: A Graph-grounded Metric for Audio Captioning
  - Lead Author: Manh Luong
- MAPLE: Multi-Agent Adaptive Planning with Long-Term Memory for Table Reasoning
  - Lead Author: Ye Bai
- Thinker-DDM: Modeling Deliberation for Machine Translation with a Drift-Diffusion Process
  - Lead Author: Hongbin Na
- Understanding Multilingual ASR Systems: The Role of Language Families and Typological Features in Seamless and Whisper
  - Lead Author: Simon Gonzalez
16:45: Session #4 - Lightning Talks (Topic: Hallucination & Vulnerabilities, Session Chair: Necva Bölücü)
- CREAM-RAG: Enhanced Retrieval Augmented Generation to Limit Hallucination through Consistency-based Self-RAG
  - Lead Author: Yuliah Louis
- GRAVITI: Grounded Retrieval Generation Framework for VideoLLM Hallucination Mitigation
  - Lead Author: Ahmad Khalil
- Privacy Beyond Memorization: Replicating and Extending Attribute-Inference Risks in LLMs
  - Lead Author: Ny Fenohery Jeannot
- Securing Off-the-Shelf LLMs against Backdoor Attacks
  - Lead Author: Weijun Li
- Who Wrote This? Attribution Across Human And Machine Authors
  - Lead Author: Anudeex Shetty

Venue: Friend in Hand Hotel - Top Bar

Location: 58 Cowper St, Glebe, NSW, 2037

Join us for dinner and drinks at a local bar, with pool tables and more. It is a 17 minute walk to the bar from the conference venue.

Friday, November 28