San Diego or Copenhagen: Meet us at NeurIPS and EurIPS 2025

The ELLIS Institute Tübingen is heading to two top-tier conferences this December: NeurIPS in San Diego and EurIPS in Copenhagen! We are excited to be part of both events, connecting scientists from around the world through outstanding research from December 2–7, 2025.

For the first time this year, the 39th annual Neural Information Processing Systems (NeurIPS) conference, one of the world’s leading events in artificial intelligence and machine learning, is supporting a new, independent NeurIPS-endorsed event in Copenhagen called EurIPS. This collaboration gives researchers the exciting opportunity to share their work at either NeurIPS or EurIPS, with papers accepted at NeurIPS optionally eligible for presentation at EurIPS as well. 

At NeurIPS 2025, our Principal Investigators will be represented with 26 contributions, including 2 talks and 6 spotlights. If you’re in San Diego this year, be sure to stop by and meet some of our PIs as they present their research (full list of contributions below).

We’re also thrilled to be a proud sponsor of EurIPS 2025, where you can find us at Booth D03-08. Stop by to say hello, learn more about our Institute’s mission and research, and explore our current open positions.

Looking forward to connecting with you at both events!

Talks
Generalized Linear Mode Connectivity for Transformers	Alexander Theus · Alessandro Cabodi · Sotiris Anagnostidis · Antonio Orvieto · Sidak Pal Singh · Valentina Boeva
In Search of Adam’s Secret Sauce	Antonio Orvieto · Robert Gower

Spotlights
The Hawthorne Effect in Reasoning Models: Evaluating and Steering Test Awareness	Sahar Abdelnabi · Ahmed Salem
OS-Harm: A Benchmark for Measuring Safety of Computer Use Agents	Thomas Kuntz · Agatha Duzan · Hao Zhao · Francesco Croce · J. Zico Kolter · Nicolas Flammarion · Maksym Andriushchenko
Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach	Jonas Geiping · Sean McLeish · Neel Jain · John Kirchenbauer · Siddharth Singh · Brian Bartoldson · Bhavya Kailkhura · Abhinav Bhatele · Tom Goldstein
TabArena: A Living Benchmark for Machine Learning on Tabular Data	Nick Erickson · Lennart Purucker · Andrej Tschalzev · David Holzmüller · Prateek Desai · David Salinas · Frank Hutter
Do-PFN: In-context Learning for Causal Effect Estimation	Jake Robertson · Arik Reuter · Siyuan Guo · Noah Hollmann · Frank Hutter · Bernhard Schölkopf
Fixed-Point RNNs: Interpolating from Diagonal to Dense	Sajad Movahedi · Felix Sarnthein · Nicola Muca Cirone · Antonio Orvieto

Posters
Contextual Integrity in LLMs via Reasoning and Reinforcement Learning	Guangchen (Eric) Lan · Huseyin A. Inan · Sahar Abdelnabi · Janardhan Kulkarni · Lukas Wutschitz · Reza Shokri · Christopher Brinton · Robert Sim
Direct Alignment with Heterogeneous Preferences	Ali Shirali · Arash Nasr-Esfahany · Abdullah Alomar · Parsa Mirtaheri · Rediet Abebe · Ariel Procaccia
Quantifying Uncertainty in Error Consistency: Towards Reliable Behavioral Comparison of Classifiers	Thomas Klein · Sascha Meyen · Wieland Brendel · Felix A. Wichmann · Kristof Meding
Reparameterized LLM Training via Orthogonal Equivalence Transformation	Zeju Qiu · Simon Buchholz · Tim Xiao · Maximilian Dax · Bernhard Schölkopf · Weiyang Liu
EquiTabPFN: A Target-Permutation Equivariant Prior Fitted Network	Michael Arbel · David Salinas · Frank Hutter
Gompertz Linear Units: Leveraging Asymmetry for Enhanced Learning Dynamics	Indrashis Das · Mahmoud Safari · Steven Adriaensen · Frank Hutter
Learning in Compact Spaces with Approximately Normalized Transformer	Jörg Franke · Urs Spiegelhalter · Marianna Nezhurina · Jenia Jitsev · Frank Hutter · Michael Hefenbrock
DeltaProduct: Improving State-Tracking in Linear RNNs via Householder Products	Julien Siems · Timur Carstensen · Arber Zela · Frank Hutter · Massimiliano Pontil · Riccardo Grazzi
GPAS: Accelerating Convergence of LLM Pretraining via Gradient-Preserving Activation Scaling	Tianhao Chen · Xin Xu · Zijing Liu · Pengxiang Li · Xinyuan Song · Ajay Jaiswal · Fan Zhang · Jishan Hu · Yang Wang · Hao Chen · Shizhe Diao · Shiwei Liu · Yu Li · Lu Yin · Can Yang
The Curse of Depth in Large Language Models	Wenfang Sun · Xinyuan Song · Pengxiang Li · Lu Yin · Yefeng Zheng · Shiwei Liu
AlphaDecay: Module-wise Weight Decay for Heavy-Tailed Balancing in LLMs	Di He · Ajay Jaiswal · Songjun Tu · Li Shen · Ganzhao Yuan · Shiwei Liu · Lu Yin
Collective Reasoning in Performative Prediction	Haiqing Zhu · Tijana Zrnic · Celestine Mendler-Dünner
Performative Validity of Recourse Explanations	Gunnar König · Hidde Fokkema · Timo Freiesleben · Celestine Mendler-Dünner · Ulrike Luxburg
Enhancing Optimizer Stability: Momentum Adaptation of The NGN Step-size	Rustem Islamov · Niccolò Ajroldi · Antonio Orvieto · Aurelien Lucchi
Counterfactual reasoning: an analysis of in-context emergence	Moritz Miller · Bernhard Schölkopf · Siyuan Guo
Language Models Are Inefficient Reasoners: An Analysis on Arithmetic Proof Search	Andreas Opedal · Yanick Zengaffinen · Haruki Shirakami · Clemente Pasti · Mrinmaya Sachan · Abulhair Saparov · Ryan Cotterell · Bernhard Schölkopf
SPARTAN: A Sparse Transformer World Model Attending to What Matters	Anson Lei · Bernhard Schölkopf · Ingmar Posner
Cultural Alien Sampler: Open-ended Art Generation Balancing Originality and Coherence	Alejandro Hernandez · Hiromu Yakura · Levin Brinkmann · Mar Canet Solal · Hassan Abu Alhaija · Ignacio Serna · Nasim Rahaman · Bernhard Schölkopf · Iyad Rahwan