Select a calendar:
Filter December Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Conferences, Lectures, & Seminars
Events for December
-
Epstein Institute, ISE 651 Seminar Class _LAST CLASS for FALL SEMESTER
Tue, Dec 03, 2024 @ 03:30 AM - 04:50 PM
Daniel J. Epstein Department of Industrial and Systems Engineering
Conferences, Lectures, & Seminars
Speaker: Dr. Lu Lu, Assistant Professor, Department of Statistics and Data Science, Yale University
Talk Title: Physics-Informed Deep Learning: Blending Data and Physics for Learning Functions and Operators
Host: Dr. Qiang Huang
More Information: FLYER 651 Dr. Lu Lu 12.3.24.png
Location: Social Sciences Building (SOS) - B2
Audiences: Everyone Is Invited
Contact: Casi Jones/ ISE
-
ECE Seminar: Biologically Inspired Algorithm and Hardware Co-Design for Efficient Machine Intelligence
Wed, Dec 04, 2024 @ 10:00 AM - 11:00 AM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Dr. Priya Panda, Assistant Professor, Electrical & Computer Engineering Department, Yale University
Talk Title: Biologically Inspired Algorithm and Hardware Co-Design for Efficient Machine Intelligence
Abstract: Artificial Intelligence (AI) is poised to revolutionize society, yet its escalating energy demands pose a formidable challenge to its long-term sustainability. The staggering gap in energy consumption between biological (Human Brain @20watts) and artificial intelligence (ChatGPT @100KWatts) is striking. My research aims to bridge this gap with a bio-inspired, integrative approach, where algorithm-hardware co-design and neuromorphic computing converge to create intelligent, energy-efficient systems.
In this talk, I will talk about my group’s recent efforts towards enabling and democratizing spike-based machine intelligence design, simulation, and evaluation across different applications. I’ll explore the distinctive benefits of Spiking Neural Networks (SNNs), especially the use of temporal dynamics, which enhances robustness while offering significant gains in latency, energy efficiency, and accuracy in tasks like video segmentation, human activity recognition, and event sensing. From a hardware perspective, I’ll examine how memory and sparsity management can accelerate SNNs on general-purpose platforms, introducing techniques like input-aware dynamic temporal exit and scaling-free quantization for efficient weight and activation compression.
Finally, I will share a vision for the future of energy-efficient AI, where our ongoing efforts in input-aware adaptive computation for large foundation models hold promise for developing end-to-end edge cloud intelligent systems capable of visual, language and multi-faceted visual-language processing. This approach opens the door to deploying low-power embodied AI and robotics.
Biography: Priya Panda is an assistant professor in the Electrical & Computer Engineering department at Yale University, USA and a Visiting Faculty Researcher at Google DeepMind with the vision and compilers/architectures team. She received her B.E. and Master's degree from BITS, Pilani, India in 2013 and her Ph.D. from Purdue University, USA in 2019. During her PhD, she interned in Intel Labs where she developed large scale spiking neural network algorithms for benchmarking the Loihi chip. She is the recipient of the 2019 Amazon Research Award, 2022 Google Research Scholar Award, 2022 DARPA Riser Award, 2023 NSF CAREER Award, 2023 DARPA Young Faculty Award, and the inaugural 2024 Purdue Engineering 38 under 38 award. She has also received the 2022 ISLPED Best Paper Award, 2022 IEEE Brain Community Best Paper Award and 2024 ASP-DAC Best Paper Nomination. Her research interests lie in Spiking Neural Networks, Efficient AI algorithm and hardware design.
Host: Dr. Peter Beerel, pabeerel@usc.edu
Webcast: https://usc.zoom.us/j/96755228104?pwd=NR5BYktbr3Yw36DWAtj5cakkt1qQR0.1 (USC NetID login required)Location: 132
WebCast Link: https://usc.zoom.us/j/96755228104?pwd=NR5BYktbr3Yw36DWAtj5cakkt1qQR0.1 (USC NetID login required)
Audiences: Everyone Is Invited
Contact: Mayumi Thrasher
-
AME Seminar
Wed, Dec 04, 2024 @ 03:30 PM - 04:30 PM
Aerospace and Mechanical Engineering
Conferences, Lectures, & Seminars
Speaker: Moumita Das, Rochester Institute of Technology
Talk Title: TBD
Host: AME Department
More Info: https://ame.usc.edu/seminars/
Webcast: https://usc.zoom.us/j/96060458816?pwd=8LmoG2q6vBCQubqqWpcizd2F1bxqsH.1Location: Seaver Science Library (SSL) - 202
WebCast Link: https://usc.zoom.us/j/96060458816?pwd=8LmoG2q6vBCQubqqWpcizd2F1bxqsH.1
Audiences: Everyone Is Invited
Contact: Tessa Yao
Event Link: https://ame.usc.edu/seminars/
-
OKRA Forum: Ranjit Singh Atwal
Thu, Dec 05, 2024 @ 10:00 AM - 11:00 AM
Alfred E. Mann Department of Biomedical Engineering, USC Viterbi School of Engineering
Conferences, Lectures, & Seminars
Speaker: Ranjit Singh Atwal, Moderna Global Fellow and Research Assistant Professor in Kelley Laboratories at Northwestern University Feinberg School of Medicine
Talk Title: LeaPFrog: Highly Scalable Cell Profiling for Druggable Target Discovery and Therapeutics Development
Abstract: Registration is required for this event: https://northwestern.zoom.us/webinar/register/8017315250267/WN_6P8qnb64Qn-VQM84m3tbXA
Large-scale genetic perturbation and cell profiling technologies have revolutionized the field of molecular biology and has the potential to transform many aspects of healthcare and biotechnology in the coming decade. Nowadays, CRISPR/Cas9-based genome-scale functional genetic screens are being routinely used to identify key genetic regulators of a phenotype of interest. However, the identification of genetic modifications that lead to a phenotypic change requires sorting large numbers of cells, which increases operational times and costs and often limits cell viability. To fully realize the potential of whole-genome CRISPR screening, advances in high-throughput cell sorting technologies are needed. Over the last 5 years, our research group has developed the use of immunomagnetic cell sorting facilitated by microfluidic chips as a rapid and scalable screening platform (termed LeaPFroG) for efficiently and accurately analyzing large numbers of CRISPR-edited cells. I will present how we have leveraged the high-throughput cell sorting capabilities of our LeaPFroG platform as a discovery engine to identify and validate novel checkpoint inhibitors for modulating tumor cell/immune cell interactions and elucidating allele-specific functional regulators of previously undruggable proteins. Lastly, I will outline how the experiences and lessons from these functional studies are being applied to the identification of cellular determinants impacting the mesangial cells in IgA Nephropathy.
Biography: Dr. Ranjit Singh Atwal is a Moderna Global Fellow and Research Assistant Professor in Kelley Laboratories at Northwestern University Feinberg School of Medicine in the Department of Biochemistry and Molecular Genetics. Dr. Atwal received his Ph.D. from McMaster University (Canada) and was a postdoctoral fellow at the Center for Genomic Medicine and faculty member at Massachusetts General Hospital and Harvard Medical School. His research investigations are focused on expanding the use of large-scale phenotypic screening technologies to address unmet needs across diverse biological realms. Current research investigations include the identification of functional regulators of undruggable proteins, rare-cell enrichment based in vivo phenotypic CRISPR screening to identify genetic regulators of metastasis and the use of tissue-selective delivery systems for therapeutic genome editing applications. He is also a founding scientist of a pre-seed startup focused on translating findings at the bench towards the development of targeted therapeutics. As part of the NU-OKRA/NUGokidney Resource Development Core, he is supporting the adaptation of the LeaPFroG platform to enable rapid enrichment of disease-relevant population of kidney cells to empower functional studies to better understand the mechanisms of kidney disease.
Host: Northwestern University & University of Southern California George M. O'Brien Kidney Resource Center
More Info: https://northwestern.zoom.us/webinar/register/8017315250267/WN_6P8qnb64Qn-VQM84m3tbXA
Webcast: https://northwestern.zoom.us/webinar/register/8017315250267/WN_6P8qnb64Qn-VQM84m3tbXAMore Information: Webinar graphic - 12.05.24.png
WebCast Link: https://northwestern.zoom.us/webinar/register/8017315250267/WN_6P8qnb64Qn-VQM84m3tbXA
Audiences: Everyone Is Invited
Contact: Greta Harrison
Event Link: https://northwestern.zoom.us/webinar/register/8017315250267/WN_6P8qnb64Qn-VQM84m3tbXA
-
NL Seminar-Title TBA
Thu, Dec 05, 2024 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Zihao He, USC/ISI
Talk Title: TBA
Series: NL Seminar
Abstract: REMINDER:
Meeting hosts only admit on-line guests that they know to the Zoom meeting. Hence, you’re highly encouraged to use your USC account to sign into Zoom.
If you’re an outside visitor, please inform us at (nlg-seminar-host(at)isi.edu) to make us aware of your attendance so we can admit you. Specify if you will attend remotely or in person at least one business day prior to the event. Provide your: full name, job title and professional affiliation and arrive at least 10 minutes before the seminar begins.
If you do not have access to the 6th Floor for in-person attendance, please check in at the 10th floor main reception desk to register as a visitor and someone will escort you to the conference room location.
Join Zoom Meeting
https://usc.zoom.us/j/98068942358?pwd=MYU6jzrZIjaIYPEuIHG0C61g3BTEXB.1
Meeting ID: 980 6894 2358
Passcode: 716186
Biography: If speaker approves to be recorded for this NL Seminar talk, it will be posted on the USC/ISI YouTube page within 1-2 business days: https://www.youtube.com/user/USCISI.
Subscribe here to learn more about upcoming seminars: https://www.isi.edu/events/
For more information on the NL Seminar series and upcoming talks, please visit:
https://www.isi.edu/research-groups-nlg/nlg-seminars/
Host: Jonathan May and Katy Felkner
More Info: https://www.isi.edu/research-groups-nlg/nlg-seminars/
Webcast: https://usc.zoom.us/j/98068942358?pwd=MYU6jzrZIjaIYPEuIHG0C61g3BTEXB.1Location: Information Science Institute (ISI) - Conf Rm#689
WebCast Link: https://usc.zoom.us/j/98068942358?pwd=MYU6jzrZIjaIYPEuIHG0C61g3BTEXB.1
Audiences: Everyone Is Invited
Contact: Pete Zamar
Event Link: https://www.isi.edu/research-groups-nlg/nlg-seminars/
-
Is Data All You Need?: Large Robot Action Models and Good Old Fashioned Engineering
Thu, Dec 05, 2024 @ 03:00 PM - 05:15 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Ken Goldberg, Ph.D., William S. Floyd Distinguished Chair of Engineering - UC Berkeley
Talk Title: Is Data All You Need?: Large Robot Action Models and Good Old Fashioned Engineering
Abstract: Enthusiasm has been skyrocketing for humanoids based on recent advances in "end-to-end" large robot action models. Initial results are promising, and several collaborative efforts are underway to collect the needed demonstration data. But is data really all you need?
Although end-to-end Large Vision, Language, Action (VLA) Models have potential to generalize and reliably solve all problems in robotics, initial results have been mixed[1]. It seems likely that the size of the VLA state space and dearth of available demonstration data, combined with challenges in getting models to generalize beyond the training distribution and the inherent challenges in interpreting and debugging large models, will make it difficult for pure end-to-end systems to provide the kind of robot performance that investors expect in the near future.
In this presentation, I share my concerns about current trends in robotics, including task definition, data collection, and experimental evaluation. I propose that to reach expected performance levels, we will need "Good Old Fashioned Engineering (GOFE)" – modularity, algorithms, and metrics. I'll present MANIP[2], a modular systems architecture that can integrate learning with well-established procedural algorithmic primitives such as Inverse Kinematics, Kalman Filters, RANSAC outlier rejection, PID modules, etc. I’ll show how we are using MANIP to improve performance on robot manipulation tasks such as grasping, cable untangling, surgical suturing, motion planning, and bagging, and propose open directions for research.
Presented at:
> Stanford Robotics Seminar, 19 April, 2024 4-min video clip
> Berkeley AI Research (BAIR) Seminar, 24 April, 2024
> IEEE ICRA Workshop, Yokohama Japan, 16 May 2024
> Berkeley Sky Lab Retreat Keynote, Santa Cruz, 29 May 2024
> Amazon Lab 126, Sunnyvale, CA, 18 June 2024
> Apple Park, Cupertino, CA, 24 July 2024
> Toyota Research Lab, San Jose, CA, 31 July 2024
> ICRA@40 Keynote, Rotterdam, 23 Sept 2024
> WAFR Keynote, Chicago, 7 Oct 2024
> Univ of Southern California (USC) Computer Science Distinguished Lecture Seminar, 5 Dec 2024
[1] Nishanth J. Kumar. Will Scaling Solve Robotics? The idea of solving the biggest robotics challenges by training large models is sparking debate.. IEEE Spectrum. 28 May 2024.
[2] MANIP: A Modular Architecture for iNtegrating Iteractive Perception into Long-Horizon Robot Manipulation Systems. Justin Yu*, Tara Sadjadpour*, Abby O’Neill, Mehdi Khfifi, Lawrence Yunliang Chen, Richard Cheng, Ashwin Balakrishna, Thomas Kollar, Ken Goldberg. IEEE/RSJ International Conference on Robots and Systems (IROS), Abhu Dhabi, UAE. Oct 2024. Paper
Please RSVP by Tuesday, December 3, 2024 (5:00 p.m., PST): https://forms.gle/w1r6Yo3se3WU8Bou7
This lecture satisfies requirements for CSCI 591: Research Colloquium.
This lecture will be presented as a HYBRID presentation, but will not be recorded. Zoom details below:
https://usc.zoom.us/j/94205149719?pwd=LjETcnHLvCzyDbB6LjjxfknZaab3Dm.1
Meeting ID: 942 0514 9719 | Passcode: 400232
Biography: Ken Goldberg is William S. Floyd Distinguished Chair of Engineering at UC Berkeley and Chief Scientist of Ambi Robotics and Jacobi Robotics. Ken leads research in robotics and automation: grasping, manipulation, and learning for applications in warehouses, industry, homes, agriculture, and robot-assisted surgery. He is Professor of IEOR with appointments in EECS and Art Practice. Ken is Chair of the Berkeley AI Research (BAIR) Steering Committee (60 faculty) and is co-founder and Editor-in-Chief emeritus of the IEEE Transactions on Automation Science and Engineering (T-ASE). He has published ten US patents, over 400 refereed papers, and presented over 600 invited lectures to academic and corporate audiences.
http://goldberg.berkeley.edu
Host: USC Thomas Lord Department of Computer Science
More Info: https://forms.gle/w1r6Yo3se3WU8Bou7
Webcast: https://usc.zoom.us/j/94205149719?pwd=LjETcnHLvCzyDbB6LjjxfknZaab3Dm.1Location: Michelson Center for Convergent Bioscience (MCB) - 101
WebCast Link: https://usc.zoom.us/j/94205149719?pwd=LjETcnHLvCzyDbB6LjjxfknZaab3Dm.1
Audiences: Everyone Is Invited
Contact: USC Thomas Lord Department of Computer Science
Event Link: https://forms.gle/w1r6Yo3se3WU8Bou7
-
PhD Defense
Fri, Dec 06, 2024 @ 10:00 AM - 12:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Talk Title: Efficient and Accurate 3D FISP=MRF at 0.55 T
Abstract: Magnetic Resonance Fingerprinting (MRF) are a set of popular multiparametric quantitative MRI techniques. With the resurgence of interest in mid- and low-field MRI, such as the 0.55 T MR system in Dynamic Imaging Science Center in USC, these techniques have gained growing research and clinical tractions. At 0.55 T, a basic fast imaging with steady-state free precession (FISP)-MRF approach has been shown feasible with promising but unexplored improvements, however, also with substantial quantification biases from reference measurements and literature values. Therefore, how to perform this approach in a more Signal-to-Noise Ratio(SNR) efficiency optimized way and how to improve its quantification accuracy have become interesting research problems.
In this dissertation, I propose a more efficient and accuracy FISP-MRF approach at 0.55 T. I start with improving 0.55 T FISP-MRF SNR efficiency and the approach produces more precise results (up to 50% smaller standard deviation values) but temporarily with unaddressed biases. It includes higher readout duty cycle, constrained reconstruction and artifacts mitigation algorithms. Then, I focus on refining RF excitation designs, which helps to partially suppress the sources of bias, resulting in more accurate quantification (~75% less bias).
Biography: Zhibo Zhu is a PhD candidate in Electrical and Computer Engineering in University of Southern California, advised by Prof. Krishna S. Nayak. He received Bachelor of Science degree in Nanjing University of Post and Telecommunication in 2015 and Master of Science degree in University of Southern California in 2017. His current research interest is improved FISP-MRF at 0.55 T MRI.
Host: Krishna Nayak
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Bella Schilter
-
Alfred E.Mann Department of Biomedical Engineering - Seminar series
Fri, Dec 06, 2024 @ 11:00 AM - 12:00 PM
Alfred E. Mann Department of Biomedical Engineering
Conferences, Lectures, & Seminars
Speaker: Carla Woods, President of Mann Healthcare Partners and Board member of USC Viterbi
Talk Title: Medical Device Companies: From Idea to a Business, and Everything in Between
Abstract: When an idea is born for a product, few face the opportunities and challenges of Medical Devices. Due diligence and planning are paramount to success. Most of the success of a new idea is not the idea itself, but all that goes into driving a successful venture and understanding how to fit the product in a dynamic existing and highly regulated environment. The opportunity that ideas generate in the medical space is to help people with health issues and improve or even save lives. These opportunities also generate jobs and economic success for collaborators, investors and businesses. Getting there involves mitigating risks, overcoming obstacles, and crossing “the valley death.” This requires investment, knowhow and talent of all types coming together. This talk will walk through examples and considerations to set up the development of an idea for success!
Biography: A USC graduate in Business Administration and Entrepreneurship, Carla has been developing and marketing medical devices for over twenty years. At Advanced Bionics Corporation, she was a key executive building the company up to and through its acquisition by Boston Scientific. She began her career at Pacesetter Systems where she planned new technology applications and product needs for pacemakers. During her tenure at Advanced Bionics/Boston Scientific, she led the business development, product development, industrial design, education, clinical research and marketing for the company and its products including the Precision Spinal Cord Stimulator, the BION® microstimulator, implantable infusion pumps, and the cochlear implant. For these products she holds over 60 U.S. patents. Carla was the recipient of the Boston Scientific Patent Milestone Award and the Advanced Bionics Business Leadership Award. She was a senior executive on the company's intellectual property review board and was the shareholder representative in the Boston Scientific acquisition of Advanced Bionics. In 2007, she became the Vice President of Program Development and Strategic Planning for the Alfred Mann Foundation for Biomedical Engineering. Carla is on the Board of the USC Viterbi School of Engineering and has served on the board of the Pacific Neuroscience Institute, AMI Institutes at USC and Purdue University, the Center for Global Innovation at the USC Marshall School of Business, the National Pain Foundation and the Fulfilment Fund.
Host: Peter Yingxiao Wang- Chair of Alfred E. Mann Department of Biomedical Engineering
Location: Ronald Tutor Hall of Engineering (RTH) - 109
Audiences: Everyone Is Invited
Contact: Carla Stanard
-
MHI - Physics Joint Seminar Series, Mark Saffman, Friday, Dec. 6th at 2pm in SSL 202
Fri, Dec 06, 2024 @ 02:00 PM - 03:30 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Mark Saffman, Department of Physics, University of Wisconsin-Madison
Talk Title: Gate Model Quantum Computing with Atom Arrays
Series: MHI Physics Joint Seminar Series
Abstract: Quantum computing with neutral atom qubits has advanced rapidly with the development of large 2D arrays and high-fidelity entangling gates. We have used atomic qubits for a variational simulation of the Lipkin-Meshkov-Glick model incorporating noise mitigation techniques. The talk will provide an overview of architectural options for neutral atom qubit arrays and present new approaches for implementing nonlocal QEC codes and fast measurements, as well as progress towards photonic remote entanglement.
Biography: Mark Saffman is an experimental physicist working in the areas of atomic physics, quantum and nonlinear optics, and quantum information processing. His research team was the first to demonstrate a quantum CNOT gate for the deterministic entanglement of a pair of neutral atoms. This was done using dipole mediated interactions between highly excited Rydberg atoms. He is currently developing scalable arrays of neutral atoms for quantum computation, communication, and sensing applications. He is the Johannes Rydberg Professor of Physics at the University of Wisconsin-Madison and has been recognized with an Alfred P. Sloan fellowship, a Vilas Associate Award, the WARF Innovation Award, and is a fellow of the American Physical Society, and Optica. He has been active in professional service including two decades as an Associate Editor at the Physical Review and is the director of The Wisconsin Quantum Institute. He also serves as Chief Scientist for Quantum Information at Infleqtion, Inc.
Host: Quntao Zhuang, Eli Levinson-Falk, Jonathan Habif, Daniel Lidar, Kelly Luo, Todd Brun, Tony Levi, Stephan Haas
More Information: Mark Saffman -Dec 6.pdf
Location: Seaver Science Library (SSL) - 202
Audiences: Everyone Is Invited
Contact: Marilyn Poplawski
-
World models beyond autoregressive next state prediction
Mon, Dec 09, 2024 @ 02:00 PM - 03:00 PM
Ming Hsieh Department of Electrical and Computer Engineering, Thomas Lord Department of Computer Science, USC School of Advanced Computing
Conferences, Lectures, & Seminars
Speaker: Abhishek Gupta, Ph.D., Assistant Professor of Computer Science and Engineering, Paul G. Allen School at the University of Washington
Talk Title: World models beyond autoregressive next state prediction
Series: CSC@USC/CommNetS-MHI Seminar Series
Abstract: Learned models of system dynamics provide an appealing way of predicting the future outcomes in a system, enabling downstream usage for planning or off-policy evaluation in applications such as robotics. However, the prevalent paradigm of autoregressive, next-state prediction in learning dynamics models is challenging to scale to environments with high dimensional observations and long horizons. In this talk, I will present alternative techniques for model learning that go beyond directly predicting next states. Firstly, we will discuss a reconstruction-free class of models that go beyond next-observation prediction by learning the evolution of task-directed latent representations for high dimensional observation spaces. We will then show how this can be generalized to learning a new class of models that avoid autoregressive prediction altogether by directly modeling long-term cumulative outcomes, while remaining task agnostic. In doing so, this talk will propose alternative ways of thinking about model learning that retain the benefits of transferability and efficiency from model-based RL, while going beyond next-state prediction.
Biography: Abhishek Gupta is an assistant professor of computer science and engineering at the Paul G. Allen School at the University of Washington. Prior to joining University of Washington, he was a post-doctoral scholar at MIT, collaborating with Russ Tedrake and Pulkit Agarwal. He completed his Ph.D. at UC Berkeley working with Pieter Abbeel and Sergey Levine, building systems that can leverage reinforcement learning algorithms to solve robotics problems. He is interested in research directions that enable directly performing reinforcement learning directly in the real world — reward supervision in reinforcement learning, large scale real world data collection, learning from demonstrations, and multi-task reinforcement learning. He has also spent time at Google Brain. He is a recipient of the NDSEG and NSF graduate research fellowships, and several of his works have been presented as spotlight presentations at top-tier machine learning and robotics conferences.
Host: Erdem Biyik
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Erdem Biyik
-
NL Seminar-Harmful Speech Detection by Language Models Exhibits Gender-Queer Dialect Bias
Thu, Dec 12, 2024 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Becca Dorn, USC/ISI
Talk Title: Harmful Speech Detection by Language Models Exhibits Gender-Queer Dialect Bias
Series: NL Seminar
Abstract: REMINDER: Meeting hosts only admit on-line guests that they know to the Zoom meeting. Hence, you’re highly encouraged to use your USC account to sign into Zoom. If you’re an outside visitor, please inform us at (nlg-seminar-host(at)isi.edu) to make us aware of your attendance so we can admit you. Specify if you will attend remotely or in person at least one business day prior to the event. Provide your: full name, job title and professional affiliation and arrive at least 10 minutes before the seminar begins. If you do not have access to the 6th Floor for in-person attendance, please check in at the 10th floor main reception desk to register as a visitor and someone will escort you to the conference room location. Join Zoom Meeting https://usc.zoom.us/j/98709918457?pwd=sVnp7kgGtL42MLRYEPaGjofzrjJFHL.1 Meeting ID: 987 0991 8457 Passcode: 592675 Content moderation on social media platforms shapes the dynamics of online discourse, influencing whose voices are amplified and whose are suppressed. Recent studies have raised concerns about the fairness of content moderation practices, particularly for aggressively flagging posts from transgender and non-binary individuals as toxic. In this study, we investigate the presence of bias in harmful speech classification of gender-queer dialect online, focusing specifically on the treatment of reclaimed slurs. We introduce a novel dataset, QueerReclaimLex, based on 109 curated templates exemplifying non-derogatory uses of LGBTQ+ slurs. Dataset instances are scored by gender-queer annotators for potential harm depending on additional context about speaker identity. We systematically evaluate the performance of five off-the-shelf language models in assessing the harm of these texts and explore the effectiveness of chain-of-thought prompting to teach large language models (LLMs) to leverage author identity context. We reveal a tendency for these models to inaccurately flag texts authored by gender-queer individuals as harmful. Strikingly, across all LLMs the performance is poorest for texts that show signs of being written by individuals targeted by the featured slur (F1 ≤ 0.24). We highlight an urgent need for fairness and inclusivity in content moderation systems. By uncovering these biases, this work aims to inform the development of more equitable content moderation practices and contribute to the creation of inclusive online spaces for all users.
Biography: Rebecca Dorn is a PhD candidate at the University of Southern California's Information Science Institute where they are co-advised by Kristina Lerman and Fred Morstatter. Previously, they earned their B.S. in Computer Science at UC Santa Cruz, advised by Lise Getoor. Their research focuses on the intersection between AI fairness, natural language processing and computational social science. Lately, their focus has surrounded how NLP systems treat dialects of historically marginalized communities.
If speaker approves to be recorded for this NL Seminar talk, it will be posted on the USC/ISI YouTube page within 1-2 business days: https://www.youtube.com/user/USCISI. Subscribe here to learn more about upcoming seminars: https://www.isi.edu/events/ For more information on the NL Seminar series and upcoming talks, please visit: https://www.isi.edu/research-groups-nlg/nlg-seminars/
Host: Jonathan May and Katy Felkner
More Info: https://www.isi.edu/research-groups-nlg/nlg-seminars/
Webcast: https://usc.zoom.us/j/98709918457?pwd=sVnp7kgGtL42MLRYEPaGjofzrjJFHL.1Location: Information Science Institute (ISI) - Conf Rm#689
WebCast Link: https://usc.zoom.us/j/98709918457?pwd=sVnp7kgGtL42MLRYEPaGjofzrjJFHL.1
Audiences: Everyone Is Invited
Contact: Pete Zamar
Event Link: https://www.isi.edu/research-groups-nlg/nlg-seminars/
-
Spring 2025: PhD New Student Orientation
Wed, Dec 18, 2024 @ 09:00 AM - 11:30 AM
USC Viterbi School of Engineering
Conferences, Lectures, & Seminars
The Viterbi PhD New Student Orientation is a mandatory online information session that will introduce you to campus services, university policies, student life, and resources available to you. You will also meet current students and Viterbi advisors. We are looking forward to welcoming all our new Viterbi Doctoral students!
Location: Online Event
Audiences: Everyone Is Invited
Contact: Sandra Balbuena
Event Link: https://engage.usc.edu/viterbi/rsvp?id=400701
