Machine Learning for Language Processing

Lectures

The notes and slides will be added incrementally.
• Lecture 1: Classification with the Perceptron
• Lecture 2: Probabilistic classification
• Lecture 3: Optimization fundamentals

*************

Reading seminars

Instructions

Please select 3 papers that you would like to present in order of preference by noon on the 20th October and email your selections to [Javascript required].

There will be 2 presentations per 50 minute session. Your presentations should be about 15 minutes allowing for a further 5 minutes for questions, and 10 minutes at the end of each session for general discussion. You should summarise the paper briefly (remember everyone will have read it), explicate any parts you found difficult or innovative, and critically evaluate the work described. For your evaluation you should consider questions like: To what extent have the stated aims of the research been achieved? To what extent is the work replicable given the information provided? In what way does the work advance the state of the art?, etc. You may prepare slides and share your screen via the data projector. You should liaise with your co-presenter to decide the order in which to make presentations. You should have all slides for the session loaded onto a single laptop set up with the data projector by the beginning of each session.

All students should read all the papers and attend all sessions prepared to discuss each paper after the presentations

Assignment

You may undertake a small project using existing datasets and machine learning software, and then submit a project report. Your project should define an experiment (probably comparative) from which you are able to draw precise and definite conclusions. Do not be too ambitious and undertake an experiment so computationally intensive that you are unable to obtain results on the hardware available to you. Alternatively, you may write an essay on any aspect of the course content. Your essay topic should involve an in-depth critical evaluation of a specific machine learning technique and its application to language processing, or of a specific language processing task and machine learning techniques that have been applied to that task. Little credit will be given for summaries of papers. In both cases, your essay or report should not exceed 5000 words and will be due on the first day of Lent Term.

You should discuss and agree your essay topic or project with both Andreas and Pushkar. Write a proposal of up to 500 words outlining the topic or project giving a preliminary reading list and indicating what resources (datasets, hardware, and toolkits / packages) you plan to use, if relevant, and send it to both of us ([Javascript required] and [Javascript required]) ASAP and no later than the 21st of October (need to have it agreed by the 3rd of November). Suitable projects will need to make use of existing labelled datasets and existing machine learning tools that are distributed and documented, so that they can be completed in reasonable time.

Schedule and Reading List

5th of November: Natural Language Inference/Textual Entailment

Papers:

Bowman et al., EMNLP 2015, A large annotated corpus for learning natural language inference

Nie et al., ACL 2020, Adversarial NLI: A New Benchmark for Natural Language Understanding

10th of November: Examining our models (and their data)

Papers:

Tu et al., TACL 2020, An Empirical Study on Robustness to Spurious Correlations using Pre-trained Language Models

Kirichenko et al., ICLR 2023, Last Layer Re-Training is Sufficient for Robustness to Spurious Correlations

12th of November: Disagreements and Uncertainty

Papers:

Jiang and de Marneffe, TACL 2022, Investigating Reasons for Disagreement in Natural Language Inference

Baan et al., EMNLP 2022, Stop Measuring Calibration When Humans Disagree

17th of November: Inference in Sequence Prediction

Papers:

Stahlberg and Byrne, EMNLP 2019, On NMT Search Errors and Model Errors: Cat Got Your Tongue?

Eikema and Aziz, Coling 2020, Is MAP Decoding All You Need? The Inadequacy of the Mode in Neural Machine Translation

19th of November: Reinforcement Learning

Papers:

Rafailov et al., NeurIPS 2023, Direct Preference Optimization: Your Language Model is Secretly a Reward Model

Sullivan and Koller, ICML 2026, GRPO is Secretly a Process Reward Model

24th of November: Hallucinations/H2>

Papers:

Si et al., NAACL 2024, Large Language Models Help Humans Verify Truthfulness – Except When They Are Convincingly Wrong

Sui et al., ACL 2024, Confabulation: The Surprising Value of Large Language Model Hallucinations

26th of November: Retrieval Augmented Generation

Papers:

Lewis et al., NeurIPS 2020, Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks

Joren et al., ICLR 2025, Sufficient Context: A New Lens on Retrieval Augmented Generation Systems

30th of November: Benchmarking

Papers:

Perez et al., NeurIPS 2021, True Few-Shot Learning with Language Models

Zouhar et al., TACL 2025, How to Select Datapoints for Efficient Human Evaluation of NLG Models?