GenAI Evaluation KDD2024:

Title: Deploying Trustworthy Generative AI

Abstract: While generative AI models and applications have huge potential across different industries, their successful commercial deployment requires addressing several ethical, trustworthiness, and safety considerations. These concerns include domain-specific evaluation, hallucinations, truthfulness and grounding, safety and alignment, bias and fairness, robustness and security, privacy, unlearning, and copyright implications, calibration and confidence, and transparency. In this talk, we first motivate the need for adopting responsible AI principles when developing and deploying large language models (LLMs), text-to-image models, and other generative AI models, and provide a roadmap for thinking about responsible AI and AI observability for generative AI in practice. Focusing on real-world generative AI use cases (e.g., evaluating LLMs for robustness, security, bias, etc. especially in health AI applications and user-facing & enterprise-internal chatbot settings), we present practical solution approaches / guidelines for applying responsible AI techniques effectively and discuss lessons learned from deploying responsible AI approaches for generative AI applications in practice. This talk will be based on our KDD'24 LLM grounding and evaluation tutorial and last year's ICML/KDD/FAccT trustworthy generative AI tutorial.

Short Bio: Krishnaram Kenthapadi is the Chief Scientist, Clinical AI at Oracle Health, where he leads the AI initiatives for Clinical Digital Assistant and other Oracle Health products. Previously, as the Chief AI Officer & Chief Scientist of Fiddler AI, he led initiatives on generative AI (e.g., Fiddler Auditor, an open-source library for evaluating & red-teaming LLMs before deployment; AI safety, observability & feedback mechanisms for LLMs in production), and on AI safety, alignment, observability, and trustworthiness, as well as the technical strategy, innovation, and thought leadership for Fiddler. Prior to that, he was a Principal Scientist at Amazon AWS AI, where he led the fairness, explainability, privacy, and model understanding initiatives in the Amazon AI platform, and shaped new initiatives such as Amazon SageMaker Clarify from inception to launch. Prior to joining Amazon, he led similar efforts at the LinkedIn AI team, and served as LinkedIn’s representative in Microsoft’s AI and Ethics in Engineering and Research (AETHER) Advisory Board. Previously, he was a Researcher at Microsoft Research Silicon Valley Lab. Krishnaram received his Ph.D. in Computer Science from Stanford University in 2006. He serves regularly on the senior program committees of FAccT, KDD, WWW, WSDM, and related conferences, and co-chaired the 2014 ACM Symposium on Computing for Development. His work has been recognized through awards at NAACL, WWW, SODA, CIKM, ICML AutoML workshop, and Microsoft’s AI/ML conference (MLADS). He has published 60+ papers, with 7000+ citations and filed 150+ patents (72 granted). He has presented tutorials on trustworthy generative AI, privacy, fairness, explainable AI, model monitoring, and responsible AI at forums such as ICML, KDD, WSDM, WWW, FAccT, and AAAI, given several invited industry talks, and instructed a course on responsible AI at Stanford.

Krishnaram Kenthapadi

Chief Scientist, Clinical AI at Oracle Health

Title: NSF funding opportunities for AI researchers

Abstract: This talk will present NSF funding opportunities that could support research in the Artificial Intelligence area, in particular on aspects related to safety, accountability, transparency, and others. The talk will briefly describe programs in AI and computational biology such as Safe Learning-Enabled Systems, Formal Methods in the Field, Molecular Foundations of Biotechnology, Smart Health and Global Centers.

Short Bio: Sorin Drăghici is a program director in the Division of Information and Intelligent Systems (IIS) of the Directorate for Computer and Information Science and Engineering (CISE) at the Science Foundation (NSF). Draghici was elected a Fellow of the Institute of Electrical and Electronics Engineers (IEEE) in 2022, for contributions to the analysis of high-throughput genomics and proteomics data.[3] He has also been elected a Fellow of the Asia-Pacific Artificial Intelligence Association (AAIA). His work is focused on research in artificial intelligence, machine learning and data mining techniques applied to bioinformatics and computational biology.

Sorin Draghici

NSF

Call for Contributions

Link to the submission website: [link]

This workshop aims to serve as a pivotal platform for discussing the forefront of Generative AI trustworthiness and evaluation advancements. Generative AI models, such as Large Language Models (LLMs) and Diffusion Models have revolutionized various domains, underscoring the critical need for reliable Generative AI technologies. As these models increasingly influence decision-making processes, establishing robust evaluation metrics and methods becomes paramount. Our objective is to delve into diverse evaluation strategies to enhance Generative AI models reliability across applications. The workshop topics include, but are not limited to:

• Holistic Evaluation: Covering datasets, metrics, and methodologies
• Trustworthiness in Generative AI Models:
• Truthfulness: Counteracting misinformation, hallucination, inconsistency, sycophancy in responses, adversarial factuality.
• Ensuring Safety and Security: privacy concerns, preventing harmful and toxicity content.
• Addressing Bias and Fairness.
• Ethical Considerations: social norm alignment, compliance with values, regulations and laws.
• Privacy: privacy awareness and privacy leakage.
• Enhancing misuse resistance, explainability, and robustness.
• User-Centric Assessment.
• Multi-perspective Evaluation: Emphasizing logical reasoning, knowledge depth, problem-solving, and user alignment.
• Cross-Modal Evaluation: Integrating text, image, audio, etc.

The workshop is designed to convene researchers from the realms of machine learning, data mining, and beyond, fostering the interdisciplinary exploration into Generative AI trustworthiness and evaluation. By featuring a blend of invited talks, presentations of peer-reviewed papers, and panel discussions, this workshop aims to facilitate exchanges of insights and foster collaborations across research and industry sectors. Participants from diverse fields such as Data Mining, Machine Learning, Natural Language Processing (NLP), and Information Retrieval are encouraged to share knowl- edge, debate challenges, and explore synergies, thereby advancing the state of the art in Generative AI technologies.

Submission Guidelines

• Paper submissions are limited to 9 pages, excluding references, must be in PDF and use ACM Conference Proceeding templates (two column format).
• Additional supplemental material focused on reproducibility can be provided. Proofs, pseudo-code, and code may also be included in the supplement, which has no explicit page limit. The supplement format could be either single column or double column. The paper should be self-contained, since reviewers are not required to read the supplement.
• The Word template guideline can be found here: [link]
• The Latex/overleaf template guideline can be found here: [link]
• The submissions will be judged for quality and relevance through single-blind reviewing.
• A paper should be submitted in PDF format through EasyChair at the following link: [link]

Organizers - Amazon Teams