Robin Shing Moon Chan

PhD Student in HCI and NLP @ ETH Zürich

robin.chan@inf.ethz.ch

Google Scholar  /  Twitter  /  LinkedIn

About Me

I'm a first-year CS PhD student at the Institute for Visual Computing and the Institute for Machine Learning at ETH Zürich. I'm advised by Prof. Menna El-Assady (IVIA). Previously, I obtained a master's degree in data science from ETH Zürich, graduating with a thesis on LLM program synthesis at IBM Research Europe (ZRL).

My main research interest lies at the intersection of human-computer interaction and natural language processing. More concretely, I want to better understand how to make humans and LLMs interact more effectively. If you're interested in similar topics, message me. I'm looking for collaborators and master's students to supervise.

Personal bits: some of my favorite books: [1, 2, 3], some of my favorite movies: [1, 2, 3]. I like making music [me Christmas caroling with friends].

News
October 2nd, 2024 The paper we wrote at IBM Research Europe on API integration with LLMs was accepted at EMNLP 2024 (Industry Track).
September 26th, 2024 Our paper On Affine Homotopy between Language Encoders was accepted at NeurIPS 2024. I will be presenting it in Vancouver in December. Thanks to all co-authors!
August 12th, 2024 We gave a tutorial about the representational capacity of neural language models at ACL 2024 in Bangkok. Check it out our interactive tutorial webpage!
May 27th, 2024 Two papers I co-authored were accepted at ACL 2024!
March 1st, 2024 Started a PhD at ETH Zürich, co-advised by Prof. Menna El-Assady and Prof. Ryan Cotterell! 🎉
September 26th, 2023 Together with Katya Mirylenka, we will give a talk at Zurich-NLP about our work at IBM Research, at the ETH AI Center. RSVP here!
July 9th, 2023 Our paper on counterfactual sample generation was accepted at ACL. I will be presenting it in Toronto in a few days! Check out our blog post about the paper!
Publications
On Affine Homotopy between Language Encoders
Robin SM Chan, Reda Boumasmoud, Anej Svete, Yuxin Ren, Qipeng Guo, Zhijing Jin, Shauli Ravfogel, Mrinmaya Sachan, Bernhard Schölkopf, Mennatallah El-Assady, Ryan Cotterell
Advances in Neural Information Processing 37 (NeurIPS), 2024 | arXiv
Abstract
Pre-trained language encoders -- functions that represent text as vectors -- are an integral component of many NLP tasks. We tackle a natural question in language encoder analysis: What does it mean for two encoders to be similar? We contend that a faithful measure of similarity needs to be intrinsic, that is, task-independent, yet still be informative of extrinsic similarity -- the performance on downstream tasks. It is common to consider two encoders similar if they are homotopic, i.e., if they can be aligned through some transformation. In this spirit, we study the properties of affine alignment of language encoders and its implications on extrinsic similarity. We find that while affine alignment is fundamentally an asymmetric notion of similarity, it is still informative of extrinsic similarity. We confirm this on datasets of natural language representations. Beyond providing useful bounds on extrinsic similarity, affine intrinsic similarity also allows us to begin uncovering the structure of the space of pre-trained encoders by defining an order over them.

Language Encoders · Metric Spaces · Large Language Models

Adapting LLMs for Structured Natural Language API Integration
Robin SM Chan, Katsiaryna Mirylenka, Thomas Gschwind, Christoph Miksovic-Czasch, Paolo Scotton, Enrico Toniato, Abdel Labbi
Proceedings of EMNLP: Industry Track, 2024 | arXiv
Abstract
Integrating APIs is crucial for enterprise systems, enabling seamless application interaction within workflows. However, the vast and diverse API landscape makes combining calls based on user intent a significant challenge. Existing methods rely on Named Entity Recognition (NER) and knowledge graphs, but struggle with control flow structures like conditionals and loops. We propose a novel framework that leverages the success of Large Language Models (LLMs) in code generation for natural language API integration. Our approach involves fine-tuning an LLM on automatically generated API flows derived from services' OpenAPI specifications. This aims to surpass NER-based methods and compare the effectiveness of different tuning strategies. Specifically, we investigate the impact of enforcing syntax through constrained generation or retrieval-augmented generation. To facilitate systematic comparison, we introduce targeted test suites that assess the generalization capabilities and ability of these approaches to retain structured knowledge. We expect to observe that fine-tuned LLMs can: (a) learn structural constraints implicitly during training, and (b) achieve significant improvements in both in-distribution and out-of-distribution performance.

Large Language Models · LM Training · API Integration

What Languages are Easy to Language-Model? A Perspective from Learning Probabilistic Regular Languages
Nadav Borenstein, Anej Svete, Robin SM Chan, Josef Valvoda, Franz Nowak, Isabelle Augenstein, Eleanor Chodroff, Ryan Cotterell
Proceedings of ACL, 2024 | arXiv
Abstract
What can large language models learn? By definition, language models (LM) are distributions over strings. Therefore, an intuitive way of addressing the above question is to formalize it as a matter of learnability of classes of distributions over strings. While prior work in this direction focused on assessing the theoretical limits, in contrast, we seek to understand the empirical learnability. Unlike prior empirical work, we evaluate neural LMs on their home turf—learning probabilistic languages—rather than as classifiers of formal languages. In particular, we investigate the learnability of regular LMs (RLMs) by RNN and Transformer LMs. We empirically test the learnability of RLMs as a function of various complexity parameters of the RLM and the hidden state size of the neural LM. We find that the RLM rank, which corresponds to the size of linear space spanned by the logits of its conditional distributions, and the expected length of sampled strings are strong and significant predictors of learnability for both RNNs and Transformers. Several other predictors also reach significance, but with differing patterns between RNNs and Transformers.

Large Language Models · Formal Language Theory

On Efficiently Representing Regular Languages as RNNs
Anej Svete, Robin SM Chan, Ryan Cotterell
Findings of ACL, 2024 | arXiv
Abstract
Recent work by Hewitt et al. (2020) provides a possible interpretation of the empirical success of recurrent neural networks (RNNs) as language models (LMs). It shows that RNNs can efficiently represent bounded hierarchical structures that are prevalent in human language. This suggests that RNNs' success might be linked to their ability to model hierarchy. However, a closer inspection of Hewitt et al.'s (2020) construction shows that it is not limited to hierarchical LMs, posing the question of what other classes of LMs can be efficiently represented by RNNs. To this end, we generalize their construction to show that RNNs can efficiently represent a larger class of LMs: Those that can be represented by a pushdown automaton with a bounded stack and a generalized stack update function. This is analogous to an automaton that keeps a memory of a fixed number of symbols and updates the memory with a simple update mechanism. Altogether, the efficiency of representing this diverse class of LMs with RNN LMs suggests novel interpretations of their inductive bias.

Language Model Expressivity · Formal Language Theory

Which Spurious Correlations Impact Reasoning in NLI Models? A Visual Interactive Diagnosis through Data-Constrained Counterfactuals
Robin Chan, Afra Amini, Mennatallah El-Assady
Proceedings of ACL: System Demonstrations, 2023 |  arXiv | blog post
Abstract
We present a human-in-the-loop dashboard tailored to diagnosing potential spurious features that NLI models rely on for predictions. The dashboard enables users to generate diverse and challenging examples by drawing inspiration from GPT-3 suggestions. Additionally, users can receive feedback from a trained NLI model on how challenging the newly created example is and make refinements based on the feedback. Through our investigation, we discover several categories of spurious correlations that impact the reasoning of NLI models, which we group into three categories: Semantic Relevance, Logical Fallacies, and Bias. Based on our findings, we identify and describe various research opportunities, including diversifying training data and assessing NLI models' robustness by creating adversarial test suites.

Counterfactuals · Mixed-Initiative Learning · Language Model Biases


Website source. Consider using Leonid Keselman's Jekyll fork of this page.