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 |