MIND (Mixed-Initiative Next-gen Design) Workshop

Workshop @ IUI'2025
March 24th, 2025

Cagliari, Italy

The explosion of Large Language Models (LLMs), particularly auto-regressive as agents serving as chatbots, generative search and work automating tools, has also brought with it inherent limitations.

We posit that efforts to alleviate these LLM challenges – hallucinations, unpredictability outputs, lack of transparency and difficulties in customization cannot be solved through algorithmic improvements alone, but requires elevated mixed-initiative interface design that is at the heart of the IUI community.

This workshop provides a forum for discussing, brainstorming, and prototyping next-generation interfaces that leverage the strengths of both agent-based and direct manipulation paradigms, ultimately empowering end-users to harness the power of LLMs more efficiently and confidently.

Workshop Goals

Start: 2025-03-24T09:00:00.000Z
End: 2025-03-24T17:45:00.000Z

The central theme of this workshop revolves around how to combine agent-based approaches and direct manipulation interfaces, two seemingly disparate paradigms. How can the strengths of both be leveraged to empower users? What does the ideal balance look like for different user types? What novel design guidelines can future practitioners adhere to when designing human-AI interaction? By bringing together researchers and practitioners from HCI, IUI, and AI, we aim to establish a joint community that can create a deeper understanding of next-generation interactive AI systems and the creation of delightful, empowering, and effective user experiences.

Key Date

Paper, Demo, & Poster Submission
January 31st, 2025 AoE
Acceptance Notifcations
February 19th, 2025
Camera-Ready Submissions
February 27th, 2025

Topics

We invite researchers, designers, engineers, and HCI practitioners from the IUI community and beyond who are interested in pushing the boundaries of interactive AI systems. We particularly welcome contributions that explore innovative approaches combining agent-based interfaces with direct manipulation techniques. This includes, but is not limited to, work in areas such as:

• Mixed-Initiative Interaction Design for LLMs • Explainable and Transparent AI (XAI) for LLMs • Human-AI Collaboration and Control in Creative Tasks • Adaptive and Personalized Interfaces for AI Systems • Evaluation of Mixed-Initiative LLM Interfaces

Submissions

Papers & Demos
Please format your paper or demo submission using the CEUR single-column template. The submission must be a minimum of 5 pages and a maximum of 10 pages for publication in CEUR.
Please submit through this link:
Submission Link

Paper Submission Template

CEUR Template

Poster
We also welcome submissions of two-page extended abstracts for our poster session. Please note that these abstracts will not be included in the CEUR proceedings. Formatting should follow the same CEUR guidelines as outlined above.

Workshop Agenda

24 Mar

A Pilot Study: Naive smart interfaces can cause accidents - Christian Arnold, Paul Robertson, Zoe Robertson, Robert Laddaga, Boris Katz and Andrei Barbu SecureLLM: New private and confidential interfaces with LLMs - Abdulrahman Alabdulkareem, Christian Arnold, Yerim Lee, Pieter M Feenstra, Boris Katz and Andrei Barbu

TextVision: A more efficient way to work with research - Melis Aslan, Maximilian Bosse, Daniel Ehlers, Marlon Hinz, Philipp Olschewski, Jannik Podszun, Elias Scharlach, Leon Selzer, Yukun Wu, Aliki Anagnostopoulou, and Daniel Sonntag

CalPal: An Intelligent Multimodal Digital Wall Calendar - Devin Murphy, Jenny Moralejo, and Paul Pu Liang Generative AI’s aggregated knowledge versus web-based curated knowledge - Ted Selker, Yunzi Wu