Banner Banner

ICLR 2025 Conference Contributions

BIFOLD at ICLR 2025

The BIFOLD research groups Machine Learning and Explainable Machine Learning in Medicine will participate in ICLR 2025, contributing eight papers, one blog article, and one talk. The Thirteenth International Conference on Learning Representations (ICLR 2025) is scheduled for April 24 – April 28 in Singapore.

The ICLR is globally renowned for presenting and publishing cutting-edge research on all aspects of deep learning used in artificial intelligence, statistics, and data science, as well as critical application areas such as machine vision, computational biology, speech recognition, text understanding, gaming, and robotics.

Below, we provide a list of contributions sorted by research group:

  • Talk: Explainable AI for Unsupervised Learning:Turning Raw Data into Scientific Insights
    By: Grégoire Montavon
    Event: ICLR 2025 Workshop: XAI4Science

  • Title: Navigating Neural Space: Revisiting Concept Activation with Pattern-Based Probes
    Authors: Frederik Pahde, Maximilian Dreyer, Moritz Weckbecker, Leander Weber, Christopher J. Anders, Thomas Wiegand, Wojciech Samek, Sebastian Lapuschkin
    Type: Main Conference Paper
    DOI: Link
    Code: GitHub

  • Title: Cat, Rat, Meow: On the Alignment of Language Models with Representations of Visual Brains
    Authors: Lorenz Linhardt, Tom Neuhäuser, Lenka Tětková, Wieland Brendel
    Presenter: Lorenz Linhardt
    Type: Poster, Workshop Paper — ICLR 2025 Workshop on Representational Alignment
    DOI: Link

  • Title: Latent Diffusion U-Net Representations Contain Intrinsic Geometry
    Authors: Jonas Loos, Lorenz Linhardt
    Presenter: Jonas Loos 
    Type: Poster, Workshop Paper — ICLR 2025 Workshop on Deep Generative Model in Healthcare
    DOI: Link

  • Title: Efficient and Flexible Neural Network Training on Large Graphs via Graph Corelets
    Authors: Leander Weber, Jim Berend, Moritz Weckbecker, Alexander Binder, Thomas Wiegand, Wojciech Samek, Sebastian Lapuschkin
    Type: Workshop Paper — ICLR 2025 Workshop: XAI4Science
    DOI: Link

  • Title: Fractional Brownian Bridges for Aligned Data
    Authors: Gabriel Nobis, Arina Belova, Maximilian Springenberg, Rembert Daems, Christoph Knochenhauer, Manfred Opper, Tolga Birdal, Wojciech Samek
    Type: Workshop Paper— ICLR 2025 Workshop on Learning Meaningful Representation
    DOI: Link

  • Title: Beyond Scalars: Concept-Based Alignment Analysis in Vision Transformers
    Authors: Johanna Vielhaben, Dilyara Bareeva, Jim Berend, Wojciech Samek, Nils Strodthoff
    Type: Workshop Paper — ICLR 2025 Workshop on Representational Alignment (Re-Align)
    DOI: Link

  • Title: ASIDE: Architectural Separation of Instructions and Data in Language Models
    Authors: Egor Zverev, Evgenii Kortukov, Alexander Panfilov, Soroush Tabesh, Sebastian Lapuschkin, Wojciech Samek, Christoph H. Lampert
    Type: Workshop Paper — ICLR 2025 Workshop on Building Trust in LLMs
    DOI: Link

  • Title: Multilevel Generative Samplers for Investigating Critical Phenomena
    Authors: Ankur Singha, Elia Cellini, Kim A. Nicoli, Karl Jansen, Stefan Kühn, Shinichi Nakajima
    Presenter: Ankur Singha, Kim Nicoli
    Type: Poster
    DOILink

  • Title: An Illustrated Guide to Automatic Sparse Differentiation
    Authors: Adrian Hill, Guillaume Dalle, Alexis Montoison
    Type: Blog Article
    DOILink
    Code: Link

  • Lukas Muttenthaler supports the event as Co-Organizer in "ICLR 2025 Workshop on Representational Alignment". More Info