Final Call for Papers: ICML-2026 Workshop on Decision-Making from Offline Datasets to Online Adaptation

[Janardhan Rao (Jana) Doppa]

Dear all,

This is the final call for research papers and position papers for our ICML 2026 workshop on Decision-Making from Offline Datasets to Online Adaptation.

This workshop aims to explore methods for learning policies, acquisition strategies, and decision rules entirely from previously collected data (offline) or with a small amount of new real-world data (online), spanning settings such as black-box optimization, contextual bandits, reinforcement learning (RL), and their synergies. The workshop will highlight both foundational advances and real-world applications in domains where online experimentation is costly, unsafe, or infeasible, including scientific discovery, engineering design, healthcare, education, recommender systems, and beyond.

Important Information:

• Website: https://decision-making-offline2online-icml2026.github.io/
• Workshop date: July 10/11, 2026
• Location: Seoul, South Korea
• Submission deadline: May 7th, 2026 AOE
• Submission site: https://openreview.net/group?id=ICML.cc/2026/Workshop/DEMO

Keynote Speakers

• Jake Gardner (University of Pennsylvania)
• Wen Sun (Cornell University)
  
• Clara Wong-Fannjiang (Genentech)
  
• Eytan Bakshy (Meta Research)
  
• Aarti Singh (Carnegie Mellon University)
  
• Sergey Levine (UC Berkeley)
  

Topics of Interest

Topics of interest include, but are not limited to:

• Offline RL: Algorithms, theory, and applications of RL trained from offline datasets, including long-horizon and safety-constrained settings.
• Offline RL for Foundation Models: RLHF, reasoning model training, and alignment using offline data.
  
• Black-Box Optimization from Offline Data: Model-based optimization and high-throughput experimental design in few- or single-round settings.
  
• Contextual Bandits from Logged Data: Learning and evaluation using large-scale interaction logs.
  
• Off-Policy Evaluation and Policy Comparison: Reliable evaluation, confidence estimation, and counterfactual reasoning.
  
• Hybrid Offline-to-Online Learning: Methods combining offline datasets with limited online interaction.
  
• Uncertainty Quantification for Offline Decision-Making: Conformal prediction and risk-aware learning.
  
• Causal Inference from Observational Data: Leveraging causal structure for improved decision-making.
  
• Generative Models for Decision-Making: Deep generative approaches for policy learning and design optimization.
  
• Multi-Task and Multi-Objective Learning: Scaling offline methods across tasks and objectives.
  
• Benchmarks and Evaluation Protocols: Realistic datasets and metrics reflecting real-world deployment challenges.
  
• Applications in Science and Engineering: Materials discovery, drug design, chip design, robotics, healthcare, education, and industrial systems.
  

Submission Types

Full Papers: Up to 9 pages in ICML or NeurIPS format, describing mature research contributions with thorough empirical or theoretical analysis.

Short Papers: 2-4 pages in ICML or NeurIPS format, presenting preliminary results, novel ideas, or position papers (including demos, code, or benchmarks).

Organization

• Aryan Deshwal (University of Minnesota, Twin-Cities)
• Haruka Kiyohara (Cornell University)
• Willie Neiswanger (University of Southern California)
• Nghia Hoang (Washington State University)
• Syrine Belakaria (Stanford University)
• Thanh Nguyen-Tang (New Jersey Institute of Technology)
• Jana Doppa (Washington State University)

ReplyReply allForward

You can't react with an emoji to a group