Taming Large Language Models @ SIGDIAL 2023 & INLG 2023

JOIN US on Discord

The goal of our workshop is to provide a common ground for experts working on different models and tasks related to controllable text generation. The interplay between research on controlled-generation and instruction-based learning paradigms sits at the heart of our discussions. Here, they can share task-agnostic solutions, engage in evaluations, and discuss task-specific challenges. By encouraging communication, collaboration, and research amongst attendees, the workshop intends to foster insightful debates about the recent progress of the NLG community.

As we witness the rapid evolution of the NLG landscape, the advent of powerful generative models promises to herald a new era across various industries. These models carry the potential to replace a significant portion of the workforce across diverse job roles. However, alongside the excitement, this transition also brings along the critical task of ensuring the controllability of these systems. Controllable NLG is not only crucial for practical applications like editing assistants, creative AI, and persona-based dialogue agents, but it is also instrumental in addressing ethical concerns, such as the potential generation of toxic, gendered, or racial content.

We cordially invite submissions for papers that contribute to the fields of controlled language generation. Whether it's an improvement on a model, a detailed analysis of a problem, a comprehensive survey, or a description of a system demo, we welcome it all. If you have a report on training a large language model on your dataset, we're particularly interested in hearing about such practical system reports.

The proceedings of the workshop will be published by the ACL Anthology. We eagerly anticipate your valuable contributions and look forward to seeing you there!

Topics of interest include, but are not limited to:

  • Alignment: Investigating techniques to better align LLMs with human values and intentions, including reward modeling, human-in-the-loop systems, and quantifying alignment metrics. Understanding the objectives pursued by a model and aligning them with human preferences are key challenges. We encourage research on methods to increase alignments, such as through prompt design and fine-tuning.
  • In-context Learning: Exploring the role of context in LLMs, including how to improve context understanding, manage context drift, and enhance context-aware responses. Also, investigating the use of in-context learning as a control mechanism.
  • Instruction-based Control: Comparing popular controlling mechanisms, including approaches such as logit manipulation, decoder mixing, and classifier guidance, amongst others, against the simpler instruction-based control.
  • Generality: Investigating controllable techniques that work cross-task and dataset.
  • Safety and Robustness: Assessing potential risks and vulnerabilities in LLMs, along with solutions such as adversarial training, safe exploration, and monitoring model behavior during deployment.
  • Controllability vs. Robustness: Developing methods to better understand LLMs' decision-making processes, and how it acts in grounded scenarios. Understanding its reliance on implicit vs. explicit memory.
  • Scalability and efficiency: Investigating novel approaches for reducing computational requirements for achieving control in LLMs.
  • Real-world applications and case studies: Showcasing successful LLM deployments in various fields, such as healthcare, finance, education, and creative industries, along with lessons learned and future opportunities.

The TamingLLM Workshop will be co-located with SIGDial 2023 and INLG 2023! We look forward to your participation in this exciting workshop!


The SIGDial-INLG 2023 conference will take place in Prague, Czech Republic in OREA Hotel Pyramida, Prague.

The hotel is a 20-minute tram ride away from the city center, within walking distance of the Prague Castle.


OREA Hotel Pyramida, Bělohorská 24, 169 01 Praha 6, Czechia.


Call for Papers

Key Date:
  • Submission deadline: July 7, 2023
  • Author notification: July 21, 2023
  • Camera-ready deadline: August 14, 2023
  • Workshop date: September 12, 2023
    All deadlines are 11.59 pm AOE time.
Submission Portal: https://softconf.com/n/tllm2023
We are pleased to announce a unique opportunity for authors who submitted to the SIGDIAL and INLG 2023 conferences. We have decided to consider papers that were reviewed but not accepted at SIGDIAL/INLG due to space constraints. If you are one of these authors and wish to be part of our "Workshop on Taming Large Language Models", we encourage you to resubmit your work to us. Your submissions will be considered in conjunction with the original anonymous reviews from SIGDIAL/INLG. To submit your paper, please directly send an email to xiangru.tang@yale.edu and include your paper. The deadline for this submission opportunity is July 21, 2023.
We welcome reports of original research in the form of two types:
  • Long papers (8 pages + references)
  • Short papers (4 pages + references)
  • The proceedings will be published by ACL Anthology.
  • All long, short, and abstract submissions must follow the two-column ACL format, which is available as an Overleaf template and also downloadable directly (Latex and Word). Please refer to the SIGDIAL 2023 website for the most recent version of the templates.
  • Submissions must conform to the official ACL style guidelines, which are contained in these templates. Submissions must be electronic, in PDF format.
  • All submissions should be anonymized to facilitate double-blind reviewing.
  • Submissions that do not adhere to the author guidelines or ACL policies will be rejected without review.
  • Appendix should be added to the main document after references.
  • Appendix does not count towards the page length.


Keynote Speakers

Shafiq Joty

Research Director @ Salesforce AI & Associate Professor @ NTU

Daphne Ippolito

Assistant Professor @ CMU

Nancy F. Chen

Senior Principal Scientist, Principal Investigator, Group Leader @ I2R, A*STAR

Nanyun (Violet) Peng

Assistant Professor @ UCLA

Eric Malmi

Research Scientist @ Google Zürich

Yuandong Tian

Research Scientist @ Meta AI Research

Nazneen Fatema Rajani

Robustness Research Lead @ Hugging Face

Accepted Papers

Accepted Papers

CST5: Data Augmentation for Code-Switched Semantic Parsing, Anmol Agarwal, Jigar Gupta, Rahul Goel, Shyam Upadhyay, Pankaj Joshi and Regarajan Aravamudhan
PandaGPT: One Model To Instruction-Follow Them All, Yixuan Su, Tian Lan, Huayang Li, Jialu Xu, Yan Wang and Deng Cai
Emotion-Conditioned Text Generation through Automatic Prompt Optimization, Yarik Menchaca Resendiz and Roman Klinger
Mitigating Harms of LLMs via Knowledge Distillation for a Virtual Museum Tour Guide, Ashley Lewis and Michael White
Evaluating Large Language Models for Document-grounded Response Generation in Information-Seeking Dialogues, Norbert Braunschweiler, Rama Sanand Doddipatla, Simon Keizer and Svetlana Stoyanchev
Enhancing Pipeline-Based Conversational Agents with Large Language Models, Mina Foosherian, Hendrik Purwins, Purna Rathnayake, Touhidul Alam, Rui Teimao and Klaus-Dieter Thoben
Style Locality for Controllable Generation with kNN Language Models, Gilles Nawezi, Lucie Flek and Charles Welch


Workshop Schedule

12 Sep 2023

Time in Prague PST Time in China
09:30-12:00 00:30-03:00 15:30-18:00 Keynote Session I
09:30-10:00 00:30-01:00 15:30-16:00 Shafiq Joty
10:00-10:30 01:00-01:30 16:00-16:30 Daphne Ippolito
10:30-11:00 01:30-02:00 16:30-17:00 Nancy F. Chen
11:00-11:30 02:00-02:30 17:00-17:30 Break (coffee provided)
11:30-12:00 02:30-03:00 17:30-18:00 Nanyun (Violet) Peng
12:00-13:45 03:00-04:45 18:00-19:45 Lunch Break
13:45-15:30 04:45-06:30 19:45-21:30 Oral Session
13:45-14:00 04:45-05:00 19:45-20:00 Style Locality for Controllable Generation with kNN Language Models
Gilles Nawezi, Lucie Flek, Charles Welch
14:00-14:15 05:00-05:15 20:00-20:15 CST5: Data Augmentation for Code-Switched Semantic Parsing
Anmol Agarwal, Jigar Gupta, Rahul Goel, Shyam Upadhyay, Pankaj Joshi, Rengarajan Aravamudhan
14:15-14:30 05:15-05:30 20:15-20:30 PandaGPT: One Model To Instruction-Follow Them All
Yixuan Su, Tian Lan, Huayang Li, Jialu Xu, Yan Wang, Deng Cai
14:30-14:45 05:30-05:45 20:30-20:45 Emotion-Conditioned Text Generation through Automatic Prompt Optimization
Yarik Menchaca Resendiz, Roman Klinger
14:45-15:00 05:45-06:00 20:45-21:00 Mitigating Harms of LLMs via Knowledge Distillation for a Virtual Museum Tour Guide
Ashley Lewis, Michael White
15:00-15:15 06:00-06:15 21:00-21:15 Evaluating Large Language Models for Document-grounded Response Generation in Information-Seeking Dialogues
Norbert Braunschweiler, Rama Doddipatla, Simon Keizer, Svetlana Stoyanchev
15:15-15:30 06:15-06:30 21:15-21:30 Enhancing Pipeline-Based Conversational Agents with Large Language Models
Mina Foosherian, Hendrik Purwins, Purna Rathnayake, Touhidul Alam, Rui Teimao, Klaus-Dieter Thoben
15:30-16:30 06:30-07:30 21:30-22:30 Break (coffee provided)
16:30-18:00 07:30-09:00 22:30-24:00 Keynote Session II
16:30-17:00 07:30-08:00 22:30-23:00 Eric Malmi
17:00-17:30 08:00-08:30 23:00-23:30 Yuandong Tian
17:30-18:00 08:30-09:00 23:30-24:00 Nazneen Fatema Rajani

We accommodate hybrid presentations: the workshop rooms will include a camera & audio equipment.
Zoom link: ZOOM (https://yale.zoom.us/my/yale.cs).
We welcome anyone to be a remote participant via Zoom and virtual registration is free (which will be available formally after August 10).


Workshop Organizers

Organizing Commitee

Xiangru Robert Tang

Ph.D. student @ Yale University

Devamanyu Hazarika

Applied Scientist @ Amazon

Di Jin

Senior Applied Scientist @ Amazon

Chao-Wei Huang

PhD Student @ National Taiwan University

Sherry Tongshuang Wu

Assistant Professor @ CMU

Advising Commitee

Bill Yuchen Lin

Postdoctoral Young Investigator @ AI2

Deng Cai

Senior Researcher @ Tencent AI Lab

Arman Cohan

Assistant Professor @ Yale University

Asli Celikyilmaz

Senior Research Manager @ Meta

Dilek Hakkani-Tür

Senior Principal Scientist @ Amazon

Publicity Committee

Libo Qin

Professor @ Central South University

Chunlin Lu

M.sc.student @ Central South University

Jingxuan Zhou

M.sc.student @ Central South University

Program Committee

Program Committee

Zhiruo Wang, Carnegie Mellon University
Guo Zhang, Massachusetts Institute of Technology
Prakhar Gupta, Carnegie Mellon University
Yilun Zhao, Yale University
Yichen Jiang, University of North Carolina at Chapel Hill
Minghao Guo, Zhejiang University
Yiming Zong, Zhejiang University
Yanjun Shao, Fudan University
Ziming Li, Hong Kong University of Science and Technology
Yuliang Liu, Nanjing University
Behnam Hedayatnia, Amazon
Xiao Zhou, Hong Kong University of Science and Technology
Boru Lu, University of Washington
Sha Li, University of Illinois at Urbana-Champaign
Deepanway Ghosal, Singapore University of Technology and Design

Contact us

Email us at tamingllm-workshop@googlegroups.com or xiangru.tang@yale.edu
We prepared for our workshop channel on the SIGdial & INLG 2023 Discord server. Use the Discord invite link https://discord.gg/zxquHwKy5 to join the Discord server.