MQU Machine Learning Reading Group

The reading group has been running weekly for several years within the Department of Computing, Macquarie University (although we only set up this github page in 2018).

We meet on Mondays by zoom. Zoom meeting details will be sent out in MLRG email announcements. As always, anyone is welcome to take part.

In a change from the last few years, our regular time is Mondays 11.00-12.30. We'll be sharing the timeslot with the LTG seminar.

Contact Mark Dras if you'd like to get more information or to join the mailing list.

Papers / Topics

2023

Date	Presenter	Topic
18/09	TC	TinyStories: How Small Can Language Models Be and Still Speak Coherent English?, Textbooks Are All You Need II: phi-1.5 technical report
14/08	MD	All That’s ‘Human’ Is Not Gold: Evaluating Human Evaluation of Generated Text
07/08	MD	Automatic Detection of Generated Text is Easiest when Humans are Fooled
07/07	MD	Bootstrapping Multilingual Semantic Parsers using Large Language Models
23/03	PF	ADePT: Auto-encoder based Differentially Private Text Transformation, When differential privacy meets NLP: The devil is in the detail
20/02	MD	Human-level play in the game of Diplomacy by combining language models with strategic reasoning
13/02	DMA	Technology behind ChatGPT: Training language models to follow instructions with human feedback, Language Models are Few-Shot Learners

2022

Date	Presenter	Topic
22/08	DW	Evolution-Guided Policy Gradient in Reinforcement Learning
01/08	MD	A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks, Distance Metric Learning with Application to Clustering with Side-Information
25/07	MD	CAPE: Context-Aware Private Embeddings for Private Language Learning
08/07	PF	Controllable Sentence Simplification, Controllable Sentence Simplification with a Unified Text-to-Text Transfer Transformer
21/03	NF	Does Label Differential Privacy Prevent Label Inference Attacks?
31/01	NF	Deep Leakage from Gradients

2021

Date	Presenter	Topic
22/11	NF	Large Language Models Can Be Strong Differentially Private Learners
11/10	MD	Optimal Geo-Indistinguishable Mechanisms for Location Privacy
23/08	PF	OptiGAN: Generative Adversarial Networks for Goal Optimized Sequence Generation
02/08	FA	Mapping Instructions to Actions in 3D Environments with Visual Goal Prediction
12/07	AD	Dense Passage Retrieval for Open-Domain Question Answering
21/06	PF	Native Language Identification with User Generated Content
07/06	ST	BERT meets Shapley: Extending SHAP Explanations to Transformer-based Classifiers
17/05	MD	CharacterBERT: Reconciling ELMo and BERT for Word-Level Open-Vocabulary Representations From Characters
19/04	TC	Are we done with ImageNet?
29/03	MD	Improving Emotion Classification through Variational Inference of Latent Variables, blog discussion
15/03	MD	Multimodal and Multi-view Models for Emotion Recognition, Improving Emotion Classification through Variational Inference of Latent Variables
22/02	YW	On the Evaluation of Vision-and-Language Navigation Instructions

2020

Date	Presenter	Topic
07/12	IW	K-Adapter: Infusing Knowledge into Pre-Trained Models with Adapters
30/11	YW	Diverse, Controllable, and Keyphrase-Aware: A Corpus and Method for News Multi-Headline Generation
09/11	AAA	UNET-Based Multi-Task Architecture for Brain Lesion Segmentation
02/11	ST	A4NT: Author Attribute Anonymity by Adversarial Training of Neural Machine Translation
26/10	--	no meeting
19/10	FA	Human Instruction-Following with Deep Reinforcement Learning via Transfer-Learning from Text
28/09	MD	Detecting Adversarial Samples Using Influence Functions and Nearest Neighbors
21/09	MD	Deep k-Nearest Neighbors: Towards Confident, Interpretable and Robust Deep Learning, Interpreting Neural Networks with Nearest Neighbors
14/09	IW	Knowledge Enhanced Contextual Word Representations
07/09	FA	Gated-Attention Architectures for Task-Oriented Language Grounding
17/08	HB	TAPAS: Weakly Supervised Table Parsing via Pre-training
10/08	MD	Privacy Risks of General-Purpose Language Models
03/08	--	postponed
27/07	AAA	Self-supervised Visual Feature Learning with Deep Neural Networks: A Survey
20/07	SS	Dynamic Routing Between Capsules
22/06	JS	Predictive Engagement: An Efficient Metric for Automatic Evaluation of Open-Domain Dialogue Systems
15/06	PJ	Improving Disfluency Detection by Self-Training a Self-Attentive Model
01/06	NL	Understanding Black-box Predictions via Influence Functions
25/05	FA	Policy Networks with Two-Stage Training for Dialogue Systems
18/05	FA	Bootstrapping a Neural Conversational Agent with Dialogue Self-Play, Crowdsourcing and On-Line Reinforcement Learning
04/05	--	postponed
06/04	NL	Assessing the Ability of LSTMs to Learn Syntax-Sensitive Dependencies
16/03	YW	Key-Value Copy Network for Novel Object Captioning
09/03	FA	Learning to Communicate with Deep Multi-Agent Reinforcement Learning
02/03	IW	OpenKI: Integrating Open Information Extraction and Knowledge Bases with Relation Inference, Matching the Blanks: Distributional Similarity for Relation Learning
24/02	TC	Adaptive neural network classifier for decoding MEG signals
17/02	CS	Heuristic Authorship Obfuscation
10/02	MD	Leveraging Hierarchical Representations for Preserving Privacy and Utility in Text
03/02	MD	Differentiation of Blackbox Combinatorial Solvers
20/01	WB	Insights into Deep Learning

2019

Date	Presenter	Topic
18/11	RS	Interpretable Machine Learning: A Guide for Making Black Box Models Explainable (cont.)
04/11	RS	Interpretable Machine Learning: A Guide for Making Black Box Models Explainable
21/10	MD	“Why Should I Trust You?” Explaining the Predictions of Any Classifier, with popular summary and notes about shortcomings
09/09	IW	Probabilistic Embedding of Knowledge Graphs with Box Lattice Measures, Smoothing the Geometry of Probabilistic Box Embeddings
26/08	-	no meeting
12/08	MD	Generating Natural Adversarial Examples (recap)
29/07	MD	A Hybrid Deep Learning Architecture for Privacy-Preserving Mobile Analytics
15/07	MD	Mask R-CNN
17/06	IW	slides on BERT
03/06	CS	Adversarially Regularized Autoencoders
20/05	MD	Understanding Convolutional Neural Networks for Text Classification
06/05	MD	Certified Robustness to Adversarial Examples with Differential Privacy
15/04	-	no meeting
01/04	MD	TEXTBUGGER: Generating Adversarial Text Against Real-world Applications
18/03	PJ	Exploring neural transducers for end-to-end speech recognition
04/03	RS	The Seven Tools of Causal Inference, with Reflections on Machine Learning, with an alternative view by Andrew Gelman
18/02	FA	Asynchronous Methods for Deep Reinforcement Learning(arxiv version)
04/02	FA	Human-level control through deep reinforcement learning
14/01	ON	SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient

2018

Date	Presenter	Topic
12/12	GK	Patient Trajectory Management: Automation of patient care prioritisation in St Vincent's Hospital ICU
05/12	-	no meeting
28/11	-	no meeting
21/11	WZ	Learning Discourse-level Diversity for Neural Dialog Models using Conditional Variational Autoencoders
14/11	MJ	Constituency Parsing with a Self-Attentive Encoder
07/11	YL	A Multilayer Convolutional Encoder-Decoder Neural Network for Grammatical Error Correction
31/10	-	no meeting
24/10	MD	Style Transfer Through Back-Translation
17/10	CS	Matching Networks for One Shot Learning
10/10	AA	Style Transfer from Non-Parallel Text by Cross-Alignment
03/10	ON	ShapeShifter: Robust Physical Adversarial Attack on Faster R-CNN Object Detector
26/09	MD	Prior Knowledge Integration for Neural Machine Translation using Posterior Regularization
19/09	MD	Visualizing and Understanding Neural Machine Translation
12/09	DM	Project Jupyter and Large Scale Data Science
05/09	JQ	An automated medical scribe for documenting clinical encounters
15/08	DM	fast.ai Deep Learning course, part 3
08/08	-	no meeting (CLaS seminar instead)
25/07	DM	fast.ai Deep Learning course, part 3
04/07	MD	fast.ai Deep Learning course, part 2 (cont.)
27/06	MD	fast.ai Deep Learning course, part 2
20/06	-	no meeting
13/06	MD	fast.ai Deep Learning course, part 1
06/06	MD	SynTF: Synthetic and Differentially Private Term Frequency Vectors for Privacy-Preserving Text Mining
30/05	MD	Towards Robust and Privacy-preserving Text Representations
23/05	CY	Stacked Convolutional and Recurrent Neural Networks for Music Emotion Recognition
16/05	MJ	Constrained RNN decoders
09/05	NF	A Novel Framework for Author Obfuscation using Generalised Differential Privacy
02/05	PA	A Scalable Hierarchical Distributed Language Model; From Softmax to Sparsemax: A Sparse Model of Attention and Multi-Label Classification
18/04	MS	An overview of deep learning based methods for unsupervised and semi-supervised anomaly detection in videos
11/04	MD	SemEval-2016 Task 6: Detecting Stance in Tweets, plus sample data; A Multidimensional Lexicon for Interpersonal Stancetaking
04/04	MD	Neural Discrete Representation Learning, plus first author's webpage
28/03	MD	Generating Natural Adversarial Examples
20/03*	PS	Variational Inference and Deep Generative Models
14/03	GK	Scalable and accurate deep learning for electronic health records
07/03	GP	An Introduction to Containers using Docker
28/02	MY	CryptoNets: Applying Neural Networks to Encrypted Data with High Throughput and Accuracy
21/02	MD	Siamese Neural Networks for One-shot Image Recognition
14/02	MJ	Key-value retrieval networks for task-oriented dialogue
07/02	MD	Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
31/01	MD	Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks

Presenters

Abbreviation	Name
AA	Abdo Alabdulwahab
AAA	Ava Assadi Abolvardi
AD	Anil Dadwal
CS	Chakaveh Saedi
CY	Chris Yong
DM	Diego Molla
DW	David Warren
FA	Fred Amouzgar
GK	Georgie Kennedy
GP	Glen Pink
HB	Hanieh Borhanazad
IW	Ian Wood
JQ	Juan Quiroz Aguilera
JS	James Simpson
MD	Mark Dras
MJ	Mark Johnson
MS	Maryam Shahpasand
MY	Mahmood Yousefiazar
NF	Natasha Fernandes
NL	Na Liu
ON	Omid Mohamad Nezami
PA	Peter Anderson
PF	Pedro Faustini
PJ	Paria Jamshid Lou
PS	Philip Schulz
RS	Rolf Schwitter
ST	Shakila Tonni
TC	Tim Chard
WB	Wray Buntine
WZ	Wei Zhang
YL	Yitao Liu
YW	Yufei Wang