Category: seminars (Page 4 of 8)

Please join us for another NLP Seminar at 11:00am in 202 South Hall on Dec 6.

Speaker: Yoav Artzi (Cornell)

Title: Robot Control and Collaboration in Situated Instruction Following


I will present two projects studying the problem of learning to follow natural language instructions. I will present new datasets, a class of interpretable models for instruction following, learning methods that combine the benefits of supervised and reinforcement learning, and new evaluation protocols. In the first part, I will discuss the task of executing natural language instructions with a robotic agent. In contrast to existing work, we do not engineer formal representations of language meaning or the robot environment. Instead, we learn to directly map raw observations and language to low-level continuous control of a quadcopter drone. In the second part, I will propose the task of learning to follow sequences of instructions in a collaborative scenario, where both the user and the system execute actions in the environment and the user controls the system using natural language. To study this problem, we build CerealBar, a multi-player 3D game where a leader instructs a follower, and both act in the environment together to accomplish complex goals.The two projects were led by Valts Blukis, Alane Suhr, and collaborators. Additional information about both projects is available here:


Yoav Artzi is an Assistant Professor in the Department of Computer Science and Cornell Tech at Cornell University. His research focuses on learning expressive models for natural language understanding, most recently in situated interactive scenarios. He received an NSF CAREER award, paper awards in EMNLP 2015, ACL 2017, and NAACL 2018, a Google Focused Research Award, and faculty awards from Google, Facebook, and Workday. Yoav holds a B.Sc. summa cum laude from Tel Aviv University and a Ph.D. from the University of Washington.

Please join us for another NLP Seminar at 4:00pm in 202 South Hall on Nov 18th. We will have two speakers visiting from Stanford.

Speaker 1: Urvashi Khandelwal

Title: Generalization through Memorization: Nearest Neighbor Language Models


Neural language models (LMs) are typically trained on large amounts of data. However, generalizing to a larger corpus or to a different domain requires additional training which is expensive. This raises an important question – how can LMs generalize better without additional training? In this talk, I will introduce kNN-LMs which extend a pre-trained LM by linearly interpolating it with a k-nearest neighbors (kNN) model. Distances are computed in the pre-trained LM embedding space, and neighbors can be drawn from any text collection, including the original LM training set. Experiments show that using the original LM training data alone, without further training, can improve performance quite a bit. In addition, kNN-LM efficiently scales up to larger training sets and allows for effective domain adaptation, by simply varying the nearest neighbor datastore, again without further training. Qualitatively, the model is particularly helpful in predicting rare patterns, such as factual knowledge. Together, these results strongly suggest that learning similarity between sequences of text is easier than predicting the next word, and that nearest neighbor search can help LMs to effectively use data without having to train on it.


Urvashi is a fifth year Computer Science PhD student at Stanford University. She works with the Stanford Natural Language Processing group, where she is advised by Prof. Dan Jurafsky. She works at the intersection of machine learning and natural language processing. More specifically, she is interested in analyzing and improving neural language models as well as sequence generation models.

Speaker 2: John Hewitt

Title: Probing Neural NLP: Ideas and Problems


Recent work in NLP has attempted to explore the basic linguistic skills induced by neural NLP models. Probing methods ask these questions through supervised analyses of models’ representations of sentences. In this talk, I’ll cover a new way of thinking about how neural networks can implicitly encode discrete structures, and provide probing evidence that ELMo and BERT have internal representations of syntax. I’ll then introduce work challenging the premises of probing, demonstrating that the methodology can admit false positive results and showing how probes can be designed and interpreted to avoid this.


John is a second year PhD student at Stanford University co-advised by Chris Manning and Percy Liang. He works on understanding the basic properties, capabilities, and limitations of neural networks for processing human language. He aims to  understand neural models for understanding’s sake, while also using the insights gained to develop models that learn and transfer more robustly from less data. He is the recipient of the EMNLP 2019 best paper runner up award.

Please join us for another NLP Seminar at 4:00pm in 202 South Hall on Oct 21st.

Speaker: Ian Tenney (Google)

Title: Probing for Structure in Sentence Representations


With the development of ELMo, BERT, and successors, pre-trained sentence encoders have become nearly ubiquitous in NLP. But what makes these models so powerful? What are they learning? A flurry of recent work – cheekily dubbed “BERTology” – seeks to analyze and explain these models, treating the encoder as an object of scientific inquiry.

In this talk, I’ll discuss a few of these analyses, focusing on our own “edge probing” work which looks at how linguistic structure is represented in deep models. Using tasks like tagging, parsing, and coreference as analysis tools, we show that language models learn strong representations of syntax but are less adept at semantic phenomena. Moreover, we find evidence of sequential reasoning, reminiscent of traditional pipelined NLP systems.

This work was jointly conducted with Patrick Xia, Berlin Chen, Alex Wang, Adam Poliak, R. Thomas McCoy, Najoung Kim, Benjamin Van Durme, Samuel R. Bowman, Dipanjan Das, and Ellie Pavlick.


Ian Tenney is a software engineer on the Language team at Google Research in Mountain View. His research focuses on understanding and analysis of deep NLP models, particularly on how they encode linguistic structure, and how unsupervised or weakly-supervised learning can give rise to complex representations and reasoning. He was a Senior Researcher on the sentence representation team for the 2018 JSALT workshop, and from 2016 to 2018 taught in the MIDS program at UC Berkeley School of Information. He holds an M.S. in Computer Science and a B.S. in Physics from Stanford.

Please join us for another NLP Seminar at 11:00 am at Soda 380 on Tuesday, Oct 8.

Speaker: Alexander Rush (Cornell)

Title: Revisiting Grammar Induction


Deep learning for NLP has become synonymous with global models trained with unlimited data. These models are incredible; however, they seem unlikely to tell us much about the way they (or language) work. Less heralded has been the ways in which deep methods have helped with inference in classical factored models. In this talk, I revisit the problem of grammar induction, an important benchmark task in NLP, using a variety of variational methods. Recent work shows that these methods greatly increase the performance of unsupervised learning methods. I argue that these approaches can be used in conjunction with global models to provide control in modern systems.


Alexander Sasha Rush is an Associate Professor at Cornell Tech. His group’s research is in the intersection of natural language processing, deep learning, and structured prediction with applications in machine translation, summarization, and text generation. He also supports open-source development including the OpenNMT project. His work has received several paper and demo awards at major NLP and visualization conferences, an NSF Career Award, and faculty awards.  He is currently the general chair of ICLR.

Please join us for another NLP Seminar at 4:00pm in 202 South Hall on Sept 30th.

Speaker: Jinfeng Rao (Facebook)

Title: Structure-Aware Learning and Decoding for Neural NLG in Task-Oriented Dialog


Generating fluent natural language responses from structured semantic representations is a critical step in task-oriented conversational systems. Previous work primarily use Seq2Seq models on flat meaning representations (MR), e.g., in the E2E NLG Challenge, which lacks of controllability of generated texts. We propose a tree-structured MR for better discourse-level structuring and sentence-level planning. We propose a constrained decoding and a tree-to-sequence approach to add structure constraints into model learning and decoding. Our experiments show both approaches lead to better semantic correctness and combining them achieves the best performance.

In addition, I will also briefly talk about my recent work on bridging the gap between relevance matching and semantic matching for short text similarity modeling.


Jinfeng Rao is currently a research scientist at Facebook Conversational AI. Before that, he was a visiting researcher at Stanford University. He obtained his PhD with Prof. Jimmy Lin from University of Maryland College Park. Jinfeng’s research interest lies at the intersection of natural language processing, information retrieval and deep learning. At Facebook, he focuses on building and shipping world-class NLG system in Assistant. He has published more than 20 articles in the major NLP/ML conferences, including ACL, EMNLP, KDD, etc. His past research helped Comcast build their XFINITY voice search system, where his proposed multi-task system has processed billions of voice queries from 20M+ voice remotes in 2019. His work also helped Comcast win the 69th Emmy Award (2017) for the technical contributions in advancing television technologies.

(Slides, available to those with email)

NLP Seminar Schedule Fall 2019

We usually meet on Mondays from 4:00-5:00pm, in  room 202 South Hall.

This is our calendar so far, and will be updated as more people are added:

Sept 23: Cristian Danescu-Niculescu-Mizil, Cornell. (talk details)

Sept 30: Jinfeng Rao, Facebook. (talk details)

Oct 8: Alexander Rush, Cornell. (talk details. time and location is different: 11 am, Tues, Soda 230)

Oct 21: Ian Tenney, Google. (talk details)

Nov 18: Urvashi Khandelwal & John Hewitt, Stanford. (talk details)

Dec 6: Yoav Artzi, Cornell. (talk details. time is different: 11 am, Fri)

If you have a email and would like to be added to the NLP seminar mailing list, email

Please join us for the first NLP Seminar of the semester on Monday, September 23,  at 4:00pm in 202 South Hall.   All are welcome!

Speaker:  Cristian Danescu-Niculescu-Mizil

Title:  Towards an artificial intuition: Conversational markers of (anti)social dynamics


Can conversational dynamics—the nature of the back and forth between people—predict outcomes of social interactions?  This talk will describe efforts on developing an artificial intuition about ongoing conversations, by modeling the subtle pragmatic and rhetorical choices of the participants.

The resulting framework distills emerging conversational patterns that can point to the nature of the social relation between interlocutors, as well as to the future trajectory of this relation.  For example, I will discuss how interactional dynamics can be used to foretell whether an online conversation will stay on track or eventually derail into personal attacks, providing community moderators several hours of prior notice before an anti-social event is likely to occur.

The data and code are available through the Cornell Conversational Analysis Toolkit (ConvoKit):

This talk includes joint work with Jonathan P. Chang, Lucas Dixon, Liye Fu, Yiqing Hua, Dan Jurafsky,  Lillian Lee, Jure Leskovec, Vlad Niculae, Chris Potts, Arthur Spirling, Dario Taraborelli, Nithum Thain, and Justine Zhang.


Cristian Danescu-Niculescu-Mizil is an assistant professor in the information science department at Cornell University.  His research aims at developing computational methods that can lead to a better understanding of our conversational practices, supporting tools that can improve the way we communicate online.  He is the recipient of several awards—including an NSF CAREER Award, the WWW 2013 Best Paper Award, a CSCW 2017 Best Paper Award, and two Google Faculty Research Awards—and his work has been featured in popular media outlets such as The Wall Street Journal, NBC’s The Today Show, NPR and the New York Times.

The NLP Seminar returns Fall semester; we will be meeting on Mondays from 4-5:30pm in 202 South Hall.

Watch this space for talk announcements.

The NLP Seminar is taking a break in Fall 2018; we won’t be scheduling speakers this semester.   See you in the Spring!

Please join us for the last NLP Seminar of the semester on Monday, April 30,  at 4:00pm in 202 South Hall.   All are welcome!

Speaker:  Marilyn Walker (UCSC)

Title:  Modeling Narrative Structure in Informal First-Person Narratives


Many genres of natural language text are narratively structured, reflecting the human bias towards organizing our experiences as narratives. Understanding narrative structure in full requires many discourse-level NLP components, including modeling the motivations, goals and desires of the protagonists, modelling the affect states of the protagonists and their transitions across story timepoints, and modelling the causal links between story events. This talk will focus on our recent work on modeling first-person participant goals and desires and their outcomes. I describe DesireDB, a collection of personal first-person stories from the Spinn3r corpus, which are annotated for statements of desire, textual evidence for desire fulfillment, and whether the stated desire is fulfilled given the evidence in the narrative context. I will describe experiments on tracking desire fulfillment using different methods, and show that a LSTM Skip-Thought model using the context both before and after the desire statement achieves an F-Measure of 0.7 on the corpus. I will also briefly discuss our work on modelling affect states and causal links between story events on the same corpus of informal stories.

The presented work was jointly conducted with Elahe Rahimtoroghi, Jiaqi Wu, Pranav Anand, Ruimin Wang, Lena Reed and Shereen Oraby.


Marilyn Walker, is a Professor of Computer Science at UC Santa Cruz, and a fellow of the Association for Computational Linguistics (ACL), in recognition of her for fundamental contributions to statistical methods for dialog optimization, to centering theory, and to expressive generation for dialog. Her current research includes work on computational models of dialogue interaction and conversational agents, analysis of affect, sarcasm and other social phenomena in social media dialogue, acquiring causal knowledge from text, conversational summarization, interactive story and narrative generation, and statistical methods for training the dialogue manager and the language generation engine for dialogue systems.

Before coming to Santa Cruz in 2009, Walker was a professor of computer science at the University of Sheffield. From 1996 to 2003, she was a principal member of the research staff at AT&T Bell Labs and AT&T Research, where she worked on the AT&T Communicator project, developing a new architecture for spoken dialogue systems and statistical methods for dialogue management and generation. Walker has published more than 200 papers and has more than 10 U.S. patents granted. She earned an M.S. in computer science at Stanford University, and a Ph.D. in computer science at the University of Pennsylvania.

« Older posts Newer posts »