[Apologies for cross-posting]
AAAI 2024 Spring Symposium on User-Aligned Assessment of Adaptive AI Systems
Stanford University, Stanford, CA, USA | March 25-27, 2024
Symposium Website
Although there is a growing need for independent assessment and regulation of AI systems, broad questions remain on the processes and technical approaches that would be required to conceptualize, express, manage, assess, and enforce such regulations for adaptive AI systems.
This symposium addresses research gaps in assessing the compliance of adaptive AI systems (systems capable of planning/learning) in the presence of post-deployment changes in requirements, in user-specific objectives, in deployment environments, and in the AI systems themselves.
These research problems go beyond the classical notions of verification and validation, where operational requirements and system specifications are available a priori. In contrast, adaptive AI systems such as household robots are expected to be designed to adapt to day-to-day changes in the requirements (which can be user-provided), environments, and as a result of system updates and learning. The symposium will feature invited talks by researchers from AI and formal methods, as well as talks on contributed papers.
Topics of interest include:
Assessment of AI system capabilities.
Algorithmic paradigms for assessment of safety and/or compliance of AI systems with evolving regulations.
Learning predictive models of agent capabilities.
Self-assessment and monitoring.
Differential assessment of AI systems following system updates or learning.
Assessment of black-box AI systems.
Types of assessment frameworks and ecosystems.
Specification languages and representations for specifying requirements on AI systems.
Assessment of LLM-based agents.
Specification and assessment of compliance w.r.t. ethics/ethical properties.
Regulation, management, and enforcement of AI assessment paradigms.
INVITED SPEAKERS (TENTATIVE)
Kamalika Chaudhuri, University of California San Diego, and Meta AI
Sriraam Natarajan, University of Texas Dallas
Stuart Russell, University of California Berkeley
Sriram Sankaranarayanan, University of Colorado Boulder
Sanjit A. Seshia, University of California Berkeley
SYMPOSIUM FORMAT
The symposium will feature invited talks, a selected set of contributed talks, and discussions. The symposium will be in-person and is scheduled for 2.5 days.
SUBMISSION INSTRUCTIONS
Submissions will take the form of extended abstracts with a first author who will be the speaker, and each speaker cannot have more than one submission as first author. Each submission must be 1-2 pages long, excluding references (in the AAAI 2024 style available here), and may refer to joint work with other collaborators. Co-authored papers should list the expected speaker as the first author of the paper.
Submissions can be a summary/survey of recent results, new research, or a position paper. There are no formal proceedings, and we encourage submissions of work presented or submitted elsewhere (no copyright transfer is required, only permission to post the abstract on the symposium site).
Papers can be submitted via EasyChair at https://easychair.org/my/conference?conf=aia2024. Additional details are available on the symposium website.
IMPORTANT DATES
Paper submission deadline: December 22, 2023 January 15, 2024 (AoE, 11:59 PM UTC-12)
Author notification: January 31, 2024
Symposium date: March 25-27, 2024
ORGANIZING COMMITTEE
Pulkit Verma, Arizona State University.
Rohan Chitnis, Meta AI.
Georgios Fainekos, Toyota Motor North America R&D.
Hazem Torfah, Chalmers University of Technology.
Siddharth Srivastava, Arizona State University.