UT NLL
UT Austin Natural Language Learning Reading Group (see past meetings before 2021 [here])
I am no longer hosting NLL! Please contact Jierui Li, the current organizer of the reading group, for more information.
I had been organizing the UT Austin Natural Language Learning Reading Group from 2023 to 2024. The schedule of our past meetings can be found here.
Fall 2024
Here's the schedule of the external talks that we are organizing this semester:- Nov 8th, 2024 Invited talk from Ryan Liu
- Dec 6th, 2024 Invited talk from Bruce W. Lee: Programming Refusal with Conditional Activation Steering
Spring 2024
- Jan 26th, 2024 Venkat will give a practice job talk.
- Feb 9th, 2024 Anubrata will give a practice job talk.
- Feb 23rd, 2024: Venkat is leading the discussion We will be reading Personas as a way to Model Truthfulness in Language Models
- March 8th, 2024 We will be reading Parallel Structures in Pre-training Data Yield In-Context Learning
- March 22nd, 2024: Juan Diego is leading the discussion We will be reading Can You Learn Semantics Through Next-Word Prediction? The Case of Entailment
- April 19th, 2024: Prof. Sasha Rush is giving a talk on State Space Models See this blog post for an overview.
Fall 2023
As usual, we will continue to focus on recent papers in NLP. At each meeting, we will have a volunteer leading the paper discussion. If you are interested in leading the paper, please let me know or simply put down your availability here.- Aug 31st, 2023: Kanishka will join our discussions on his recent outstanding paper at ACL Language model acceptability judgements are not always robust to context
- Sep 14th, 2023 Unleashing Cognitive Synergy in Large Language Models: A Task-Solving Agent through Multi-Persona Self-Collaboration
- Sep 28th, 2023: Philippe will lead the discussions Llama 2: Open Foundation and Fine-Tuned Chat Models
- Oct 12nd, 2023: Venkat will lead the discussions Scim: Intelligent Skimming Support for Scientific Papers
- Oct 26th, 2023 ELQA: A Corpus of Metalinguistic Questions and Answers about English
- Nov 9th, 2023, Nathan Schneider will give a talk on ``Toward Natural Metalanguage Processing'' [Abstract] People don't just talk with natural language: sometimes, they talk about it. A wealth of knowledge about words, grammar, and meaning is communicated metalinguistically—whether it's through dictionaries, language learning resources, scholarly works in linguistics and literature, or social/political/legal discourse. Are current NLP models fluent in metalanguage, and can they provide accurate metalinguistic explanations? I will present case studies looking at two metalinguistically rich genres: (i) online language discussion forums [ACL 2023], and (ii) judicial rulings involving language interpretation. We find that large language models can largely categorize kinds of metalanguage, and can generate satisfactory answers to some (but not all) metalinguistic questions. (Joint work with Shabnam Behzad, Michael Kranzlein, Keisuke Sakaguchi, Kevin Tobia, and Amir Zeldes.)
Outstanding Paper at ACL 2023
Spring 2023
This semester, we will focus on recent papers in NLP. At each meeting, one student should sign up for giving a short initial summary of the paper and for preparing some questions to get the discussion going. We will also allocate the last two meetings for students to talk about their own research. Students can give an update about their ongoing research in the form of a 10-min presentation. If you are interested in leading the paper or giving a presentation, please fill in your availability here. The meetings are held bi-weekly on Mondays 11:00 AM - 12:00 PM, starting from Jan 23rd. The meetings will be in hybrid, both at GDC 3.816 and via Zoom.- Jan 23rd, 2023: Hongli will lead the paper discussions Interpreting Language Models with Contrastive Explanations, EMNLP 2022
- Feb 6th, 2023: Hongli will lead the paper discussions Neural Theory-of-Mind? On the Limits of Social Intelligence in Large LMs, EMNLP 2022
- Feb 20th, 2023: Kyle will join the paper discussions Dissociating language and thought in large language models: a cognitive perspective
- Mar 6th, 2023: Venkat will lead the paper discussions The Capacity for Moral Self-Correction in Large Language Models
- Mar 20th, 2023: Juan Diego will lead the paper discussions Towards Tracing Factual Knowledge in Language Models Back to the Training Data
- April 3rd, 2023: Anubrata will lead the paper discussions Explainable AI is Dead, Long Live Explainable AI! Hypothesis-driven decision support
- April 17th, 2023: Hongli, Venkat, and Jierui will give research updates