AMI-PAKDD 2026

About the Workshop

This workshop bridges the gap between theoretical advances and practical implementations in Autonomous Machine Intelligence (AMI). Inspired by recent breakthroughs including Meta's V-JEPA and Yann LeCun's vision of autonomous intelligence, we bring together researchers and practitioners to discuss how self-supervised learning, energy-based models, and predictive architectures translate into deployable autonomous systems.

AMI directly aligns with PAKDD's Theoretical Foundations and Interdisciplinary Applications tracks. It demonstrates how autonomous systems learn from data, extract actionable knowledge, and make decisions—core themes of knowledge discovery.

Workshop Highlights

•Invited keynote from leading experts in autonomous AI
•Industry panel on deployment challenges
•Interactive poster session for theory-practice discussions
•Best paper awards for theoretical and practical contributions
•Potential live demonstrations of autonomous systems

PAKDD 2026

Hong Kong • 9–12 June

Visit PAKDD Website →

Topics of Interest

Theoretical Foundations

›Self-supervised learning and world models
›Energy-based models and latent variables
›JEPA and predictive architectures
›Contrastive learning methods

System Architectures

›Perception-planning-action cycles
›Multimodal fusion techniques
›Hierarchical representations
›Video understanding systems

Practical Applications

›Autonomous robotics and embodied AI
›Autonomous vehicles and transportation
›Industrial automation
›Medical robotics and healthcare

Accepted Papers

SCTS: Self-Critique Tree Search for Out-of-Distribution Vulnerability Detection

Zhu, Rui; Zhou, Nan; Liao, Jikang; Duan, Lixin; Yin, Guangqiang

Invariant Offline Constrained Reinforcement Learning for Circular Aquaculture under Heterogeneous Environments

Duy Le, Tan; ĐẠT, PHẠM; Nguyen, Hong Quan; Nguyen, Minh Tu; Huynh, Kha Tu

World Models as Constraints: Heterogeneous Rail Robot Coordination via CA-JEPA and LLM-Based Planning

Xue, Rui; He, Wenxuan; Zhang, Yu; Liu, Jingyuan

An Empirical Analysis of Distributional Effects in Learning-Based Routing Systems

Thyssens, Daniela; Dernedde, Tim; Schmidt-Thieme, Lars

Learn to Bid: Near-Optimal Knapsack Bidding under Budget & CPA Constraints and Sparse Delayed Feedback

Mungoli, Abhishek; Parthasarathy, Vishwath; Subramaniam, Raja; Bhamidipati, Narayan

A Benchmark for Structured Multihop Reasoning in Cross-Institution University Admission Advisory Question Answering

Nguyen, Long; Ngo, Tin; Le, Dung; Vo, Quynh; Nguyen, Dung; Le, Khang; Quan, Tho

Learning What Can Be Picked: Active Reachability Estimation for Efficient Robotic Fruit Harvesting

Syeda, Nur Afsa ; Elmahallawy, Mohamed

LatticeGuard: A Multi-Branch Adaptive Defense Framework for Graph Neural Networks

Trivedi, Abhaya; Gupta, Jhalak; Mehak

Brand Evaluation System From Big Data On The Amazon E-Commerce Platform

NGO-HO, Anh-Khoi; Nguyen, Anh-Duy; Vo, Khuong-Duy; Nguyen, Hong-Tin

Tentative Program

Date & Time

June 9, 2026

09:00 - 10:30

Venue

Tai Po IV Room, 2/F

Regal Riverside Hotel

34–36 Tai Chung Kiu Road, Sha Tin, Hong Kong

No.	Time	Paper Title	Authors
1	09:00 – 09:10	SCTS: Self-Critique Tree Search for Out-of-Distribution Vulnerability Detection	Zhu, Rui; Zhou, Nan; Liao, Jikang; Duan, Lixin; Yin, Guangqiang
2	09:10 – 09:20	Invariant Offline Constrained Reinforcement Learning for Circular Aquaculture under Heterogeneous Environments	Duy Le, Tan; ĐẠT, PHẠM; Nguyen, Hong Quan; Nguyen, Minh Tu; Huynh, Kha Tu
3	09:20 – 09:30	World Models as Constraints: Heterogeneous Rail Robot Coordination via CA-JEPA and LLM-Based Planning	Xue, Rui; He, Wenxuan; Zhang, Yu; Liu, Jingyuan
4	09:30 – 09:40	An Empirical Analysis of Distributional Effects in Learning-Based Routing Systems	Thyssens, Daniela; Dernedde, Tim; Schmidt-Thieme, Lars
5	09:40 – 09:50	Learn to Bid: Near-Optimal Knapsack Bidding under Budget & CPA Constraints and Sparse Delayed Feedback	Mungoli, Abhishek; Parthasarathy, Vishwath; Subramaniam, Raja; Bhamidipati, Narayan
6	09:50 – 10:00	A Benchmark for Structured Multihop Reasoning in Cross-Institution University Admission Advisory Question Answering	Nguyen, Long; Ngo, Tin; Le, Dung; Vo, Quynh; Nguyen, Dung; Le, Khang; Quan, Tho
7	10:00 – 10:10	Learning What Can Be Picked: Active Reachability Estimation for Efficient Robotic Fruit Harvesting	Syeda, Nur Afsa; Elmahallawy, Mohamed
8	10:10 – 10:20	LatticeGuard: A Multi-Branch Adaptive Defense Framework for Graph Neural Networks	Trivedi, Abhaya; Gupta, Jhalak; Mehak
9	10:20 – 10:30	Brand Evaluation System From Big Data On The Amazon E-Commerce Platform	NGO-HO, Anh-Khoi; Nguyen, Anh-Duy; Vo, Khuong-Duy; Nguyen, Hong-Tin