Co-located with ICSDS 2026
Location: Split, Croatia
Venue: Le Meridien Lav hotel
Date: Monday, December 14, 2026

Overview

Building on the success of last year’s inaugural launch (FSML 2025), the Institute of Mathematical Statistics (IMS) is proud to present the 2026 Frontiers in Statistical Machine Learning (FSML) workshop.

This annual series continues its mission to spotlight emerging and impactful topics in statistical machine learning that are rapidly evolving but have yet to receive significant attention in leading IMS and ASA publications. Each year, FSML focuses on specific core themes to encourage the dissemination of novel ideas and foster deeper engagement within the community.

Inspired by the dynamic format of machine learning conference workshops, FSML brings a fresh and interactive approach to the statistics landscape. The workshop hosts an open call for short paper submissions, followed by a rigorous and transparent review process. (Everyone is welcome to attend by registering — submitting is required only if you wish to present a poster.)

New for 2026: We are introducing a Fast Track submission option for papers recently accepted at major machine learning venues (e.g., NeurIPS, ICML). This addition aims to bridge the gap between communities by showcasing high-impact ML work to a broader statistical audience.

The 2026 program highlights two core themes:

  • Area 1: Generative and Foundation Models for Statistics
    • Deep Generative Modeling: Diffusion models, normalizing flows, and other generative approaches for density estimation, sampling, and synthetic data.
    • Foundation Models for Tabular Data: Pretrained and in-context models for prediction and inference on structured data.
    • LLMs for Statistical Reasoning: Large language models as tools for data analysis, hypothesis generation, and automated workflows.
    • Amortized and Simulation-Based Inference: In-context learning and pretrained networks for Bayesian and likelihood-free inference.
  • Area 2: Science of Deep Learning
    • Theoretical Foundations: Exploring mathematical and statistical principles underlying deep learning.
    • Phenomenological Studies of Learning Systems: Cataloging and explaining intriguing behaviors in learning dynamics.
    • Interpretability, Alignment, and Safety: Understanding and guiding AI systems to ensure ethical and safe operation.
    • Emerging Learning Paradigms: Investigating new approaches, such as in-context learning and scaling laws.

Keynote Speakers

Yan Shuo Tan Yan Shuo Tan National University of Singapore
Yuexi Wang Yuexi Wang University of Illinois Urbana-Champaign
Alexander Henzi Alexander Henzi Tsinghua University
Andrej Risteski Andrej Risteski Carnegie Mellon University
Pragya Sur Pragya Sur Harvard University
Yuxin Chen Yuxin Chen University of Pennsylvania

Call for Papers

We invite submissions in the two core themes listed above. FSML 2026 has two submission tracks. Both are non-archival, and accepted submissions from both are presented in person as posters. Travel awards are available for Workshop Track submissions only (see Travel Awards).

1. Workshop Track

This track is for new research ideas or work-in-progress.

  • Format: Extended abstracts of 3–5 pages (excluding references/appendices).
  • Review Process: Double-blind peer review.
  • Dual Submission:
    • Published: Previously published work must be substantially extended to be considered.
    • Under Review: Papers currently under review elsewhere may be submitted without extension (but must still confirm to 3–5 pages).
  • Submission: Submit via OpenReview.

2. Fast Track (Recent Publications)

We invite researchers to present papers on the development, analysis, or application of statistical machine learning that have previously been accepted at major machine learning conferences and journals since the inaugural FSML 2025 workshop (August 2025): NeurIPS 2025, ICLR 2026, AISTATS 2026, ICML 2026, and UAI 2026, as well as JMLR and TMLR. These papers should be formatted according to the camera-ready standards of their respective venues. Author names do not need to be anonymized. The Fast Track is non-archival and does not have proceedings, to avoid conflict with other venues’ double submission policies.

  • Format: Submit the original camera-ready PDF (with author names included).
  • Selection: Fast Track submissions are selected based on eligibility and fit with the core themes; no additional peer review is conducted.
  • Submission: Submit via Google Form.

Travel Awards

The IMS is pleased to offer $500 USD travel awards to support participation in the workshop. The top 10 applicants will receive these awards.

Eligibility:

  • Who can apply: Graduate students and postdoctoral researchers who are the first author and presenter of an accepted paper in the Workshop Track.
  • Exclusions: If you have already applied for any of the following three awards this year, you are not eligible for the FSML travel award:
    1. IMS Hannan Graduate Student Travel Award
    2. IMS New Researcher Travel Award
    3. ICSDS travel award

Selection Process: Awards are granted based on the technical quality of the submission, reviewer scores, and relevance to the core themes.

How to Apply: Please indicate your interest in a travel award during the submission process on OpenReview.

Key Dates

Submission Deadline Wednesday, September 30, 2026 (AOE)
Notification of Acceptance/Travel Award Wednesday, October 14, 2026
Registration Deadline Wednesday, November 11, 2026
Workshop Date Monday, December 14, 2026

Registration

There are three types of registration, including a special offer for students to attend our workshop:​

IMS Member US$120
Non-IMS Member US$150
Student US$60

Please click here to register.

Organizers

Contact: fsmlims@gmail.com