Skip to yearly menu bar Skip to main content


Workshop Sat, Dec 12, 2026 • 6:00 PM – 3:00 AM AEDT Salle 242

2nd Workshop on Principles of Generative Modeling (PriGM)

Francesco Cagnetta ⋅ Elisabetta Cornacchia ⋅ Valentin De Bortoli ⋅ Soon Hoe Lim ⋅ Bingbin Liu ⋅ Bruno Loureiro ⋅ Gabriel Peyré
[ OpenReview] 

Abstract

Chat is not available.

Schedule

Time zone: Conference local time zone
Free Reward Tilting via Embedding Diffusion
Rafal Karczewski ⋅ Peter Blohm ⋅ Markus Heinonen ⋅ Vikas Garg
Free Reward Tilting via Embedding Diffusion
Rafal Karczewski ⋅ Peter Blohm ⋅ Markus Heinonen ⋅ Vikas Garg
A Boltzmann Kernel for Discrete Diffusion: How Locality and Equivariance Explain Creativity
Alberto Bechara ⋅ Navonil Neogi ⋅ Nabil Iqbal ⋅ Erik Bekkers ⋅ Maria Esteban-Casadevall
A Boltzmann Kernel for Discrete Diffusion: How Locality and Equivariance Explain Creativity
Alberto Bechara ⋅ Navonil Neogi ⋅ Nabil Iqbal ⋅ Erik Bekkers ⋅ Maria Esteban-Casadevall
Tracing Generated Samples to Training-Data Clusters in Flow-Matching Models
Rania Briq ⋅ Ohad Fried ⋅ Michael Kamp ⋅ Stefan Kesselheim
Tracing Generated Samples to Training-Data Clusters in Flow-Matching Models
Rania Briq ⋅ Ohad Fried ⋅ Michael Kamp ⋅ Stefan Kesselheim
Structure over Pixels: Learning Variable-Length Visual Programs
Piotr Wyrwiński ⋅ Kacper Dobek ⋅ Krzysztof Krawiec
Structure over Pixels: Learning Variable-Length Visual Programs
Piotr Wyrwiński ⋅ Kacper Dobek ⋅ Krzysztof Krawiec
Towards More General Control of Diffusion Models Using Jeffrey Guidance
Raphaël Razafindralambo ⋅ Rémy Sun ⋅ Frederic Precioso ⋅ Jes Frellsen ⋅ Pierre-Alexandre Mattei
Towards More General Control of Diffusion Models Using Jeffrey Guidance
Raphaël Razafindralambo ⋅ Rémy Sun ⋅ Frederic Precioso ⋅ Jes Frellsen ⋅ Pierre-Alexandre Mattei
Local Coverage Governs Memorization in Diffusion Models
Claudia Lioba Merger ⋅ Sebastian Goldt
Local Coverage Governs Memorization in Diffusion Models
Claudia Lioba Merger ⋅ Sebastian Goldt
Diffusion on a Continuous Latent Space for Contextual Combinatorial Optimization
Gilles Fevry ⋅ De Castro Yohann ⋅ Axel Parmentier ⋅ Pierre-Cyril Aubin-Frankowski
Diffusion on a Continuous Latent Space for Contextual Combinatorial Optimization
Gilles Fevry ⋅ De Castro Yohann ⋅ Axel Parmentier ⋅ Pierre-Cyril Aubin-Frankowski
PeerConf: Self-Calibrating Early Stopping for Efficient Parallel LLM Reasoning
Mofe Atanda ⋅ Yani Baguenane ⋅ Tyler Yi ⋅ Deniz Filiz ⋅ Mohammed Mahfoud ⋅ Anusha Mujumdar
PeerConf: Self-Calibrating Early Stopping for Efficient Parallel LLM Reasoning
Mofe Atanda ⋅ Yani Baguenane ⋅ Tyler Yi ⋅ Deniz Filiz ⋅ Mohammed Mahfoud ⋅ Anusha Mujumdar
Generative Flow Sparse Networks
Reza Alvandi ⋅ Borna Sayedana ⋅ Pierre-Luc Bacon
Generative Flow Sparse Networks
Reza Alvandi ⋅ Borna Sayedana ⋅ Pierre-Luc Bacon
Gaussian Flow-Matching Schedules: Implications for Sampling and Training
Arsène Claustre ⋅ Hugo Negrel ⋅ Claire Boyer ⋅ Kimia Nadjahi ⋅ Eric Vanden-Eijnden
Gaussian Flow-Matching Schedules: Implications for Sampling and Training
Arsène Claustre ⋅ Hugo Negrel ⋅ Claire Boyer ⋅ Kimia Nadjahi ⋅ Eric Vanden-Eijnden
SSD: Spatially Speculative Decoding Accelerates Autoregressive Image Generation
Shilong Xiang ⋅ Zirui Zhang ⋅ Lijun Yu ⋅ Chengzhi Mao
SSD: Spatially Speculative Decoding Accelerates Autoregressive Image Generation
Shilong Xiang ⋅ Zirui Zhang ⋅ Lijun Yu ⋅ Chengzhi Mao
First-Extinction Law for Resampling Processes
Matteo Benati ⋅ Alessandro Londei ⋅ Denise Lanzieri ⋅ Vittorio Loreto
First-Extinction Law for Resampling Processes
Matteo Benati ⋅ Alessandro Londei ⋅ Denise Lanzieri ⋅ Vittorio Loreto
Limit Points of Reflow with Minibatch Optimal Transport
Antonin Chambolle ⋅ Johannes Hertrich
Limit Points of Reflow with Minibatch Optimal Transport
Antonin Chambolle ⋅ Johannes Hertrich
Iterative Value-Guided Tilting for Fine-Tuning Masked Diffusion Models: An Actor–Critic Approach
Jawad Chemaou ⋅ Le-Tuyet-Nhi PHAM ⋅ Giovanni Conforti ⋅ Eric Moulines ⋅ Yazid Janati ⋅ Alain Durmus
Iterative Value-Guided Tilting for Fine-Tuning Masked Diffusion Models: An Actor–Critic Approach
Jawad Chemaou ⋅ Le-Tuyet-Nhi PHAM ⋅ Giovanni Conforti ⋅ Eric Moulines ⋅ Yazid Janati ⋅ Alain Durmus
Mitigating Model Collapse through Dreaming Learning
Matteo Benati ⋅ Alessandro Londei ⋅ Denise Lanzieri ⋅ Raffaello Mastromarino ⋅ Vittorio Loreto
Mitigating Model Collapse through Dreaming Learning
Matteo Benati ⋅ Alessandro Londei ⋅ Denise Lanzieri ⋅ Raffaello Mastromarino ⋅ Vittorio Loreto
BiRRM: Bidirectional Token Mixing for Recurrent Reasoning Models
Leon A. Hofmeister-Akkad ⋅ Andreas Mayr ⋅ Günter Klambauer
BiRRM: Bidirectional Token Mixing for Recurrent Reasoning Models
Leon A. Hofmeister-Akkad ⋅ Andreas Mayr ⋅ Günter Klambauer
Tiny Recursive Models, Think BIGR: Boosting Inference with Guided Reasoning
Andrew Corbett ⋅ Tim Dodwell ⋅ Sai-Aakash Ramesh ⋅ Archit Sood ⋅ Anna Tzatzopoulou
Tiny Recursive Models, Think BIGR: Boosting Inference with Guided Reasoning
Andrew Corbett ⋅ Tim Dodwell ⋅ Sai-Aakash Ramesh ⋅ Archit Sood ⋅ Anna Tzatzopoulou
Improving the Exploration Behavior of Generative Recurrent Reasoning models with xRRM
Richard Freinschlag ⋅ Timo Bertram ⋅ Erich Kobler ⋅ Günter Klambauer ⋅ Andreas Mayr
Improving the Exploration Behavior of Generative Recurrent Reasoning models with xRRM
Richard Freinschlag ⋅ Timo Bertram ⋅ Erich Kobler ⋅ Günter Klambauer ⋅ Andreas Mayr
Masked Diffusion Noise Optimisation
Louis Groll ⋅ Vincent Pauline ⋅ Riccardo Barbano ⋅ Stefan Bauer
Masked Diffusion Noise Optimisation
Louis Groll ⋅ Vincent Pauline ⋅ Riccardo Barbano ⋅ Stefan Bauer
Pretraining Amortised Samplers via Unsupervised Skill Discovery
Inhyuck Song ⋅ Jinkyoo Park ⋅ Esmeralda S Whitammer ⋅ Sanghyeok Choi
Pretraining Amortised Samplers via Unsupervised Skill Discovery
Inhyuck Song ⋅ Jinkyoo Park ⋅ Esmeralda S Whitammer ⋅ Sanghyeok Choi
ParetoTransport: Generative Optimization by Mass Transport Toward The Pareto Front
Stephanie Holly ⋅ Sepp Hochreiter ⋅ Werner Zellinger
ParetoTransport: Generative Optimization by Mass Transport Toward The Pareto Front
Stephanie Holly ⋅ Sepp Hochreiter ⋅ Werner Zellinger
Do FID Evaluations Bias Generative Models Away from Statistical Relevance?
Saumya Goyal ⋅ Dhruv Garg ⋅ Divyan Goyal ⋅ Barnabas Poczos
Do FID Evaluations Bias Generative Models Away from Statistical Relevance?
Saumya Goyal ⋅ Dhruv Garg ⋅ Divyan Goyal ⋅ Barnabas Poczos
In-context classification: Calibration and noise in linear attention with softmax readout
Arman Rysmakhanov ⋅ Mary Letey ⋅ Yue Lu ⋅ Jacob Zavatone-Veth ⋅ Cengiz Pehlevan
In-context classification: Calibration and noise in linear attention with softmax readout
Arman Rysmakhanov ⋅ Mary Letey ⋅ Yue Lu ⋅ Jacob Zavatone-Veth ⋅ Cengiz Pehlevan
What Does Uniform Noise Add Beyond MASK?
Jed Guzelkabaagac ⋅ Julien Coquet ⋅ Dimitri von Rütte ⋅ Tiago Pimentel ⋅ Antonio Orvieto ⋅ Thomas Hofmann
What Does Uniform Noise Add Beyond MASK?
Jed Guzelkabaagac ⋅ Julien Coquet ⋅ Dimitri von Rütte ⋅ Tiago Pimentel ⋅ Antonio Orvieto ⋅ Thomas Hofmann
Diffusable Latents from Structure-Agnostic Distillation
Adrien RAMANANA RAHARY ⋅ Nicolas Dufour ⋅ Patrick Perez ⋅ David Picard
Diffusable Latents from Structure-Agnostic Distillation
Adrien RAMANANA RAHARY ⋅ Nicolas Dufour ⋅ Patrick Perez ⋅ David Picard
Information-Theoretic Scaling Laws for Generative Diffusion Models
Druv Pai ⋅ Dejan Stancevic ⋅ Yi Ma ⋅ Luca Ambrogioni
Information-Theoretic Scaling Laws for Generative Diffusion Models
Druv Pai ⋅ Dejan Stancevic ⋅ Yi Ma ⋅ Luca Ambrogioni
Conditioning Structure, Not Entropy Alone, Predicts Memorization Dynamics
Rodrigo Alvarez Lucendo ⋅ Madhur Panwar ⋅ Gail Weiss ⋅ Antoine Bosselut
Conditioning Structure, Not Entropy Alone, Predicts Memorization Dynamics
Rodrigo Alvarez Lucendo ⋅ Madhur Panwar ⋅ Gail Weiss ⋅ Antoine Bosselut
ENCORE: Exact Non-equilibrium COntrol with Replica Exchange for Diffusion Generation
Jiahao Yu ⋅ Saifuddin Syed ⋅ José Miguel Hernández-Lobato ⋅ Jiajun He
ENCORE: Exact Non-equilibrium COntrol with Replica Exchange for Diffusion Generation
Jiahao Yu ⋅ Saifuddin Syed ⋅ José Miguel Hernández-Lobato ⋅ Jiajun He
MoE Routing Testbed: Studying Expert Specialization and Routing Behavior at Small Scale
Tobias Falke ⋅ Nicolas Anastassacos ⋅ Samson Tan ⋅ Chankrisna R Meas ⋅ Chandana Satya Prakash ⋅ Nitesh Sekhar ⋅ M Saiful Bari ⋅ SriKrishna Kompella ⋅ Gamaleldin Elsayed
MoE Routing Testbed: Studying Expert Specialization and Routing Behavior at Small Scale
Tobias Falke ⋅ Nicolas Anastassacos ⋅ Samson Tan ⋅ Chankrisna R Meas ⋅ Chandana Satya Prakash ⋅ Nitesh Sekhar ⋅ M Saiful Bari ⋅ SriKrishna Kompella ⋅ Gamaleldin Elsayed
RREDCoT: Segment-Level Reward Redistribution for Reasoning Models
Mykyta Ielanskyi ⋅ Kajetan Schweighofer ⋅ Lukas Aichberger ⋅ Sepp Hochreiter
RREDCoT: Segment-Level Reward Redistribution for Reasoning Models
Mykyta Ielanskyi ⋅ Kajetan Schweighofer ⋅ Lukas Aichberger ⋅ Sepp Hochreiter
Minimum Stability with Noise-Injected Inputs
Yu-Han Wu ⋅ Pierre Marion
Minimum Stability with Noise-Injected Inputs
Yu-Han Wu ⋅ Pierre Marion
RUM: Reinforcement Unlearning for 3D Molecular Flow Matching Models
Lianghong Chen ⋅ Ganlin Feng ⋅ Pingzhao Hu
RUM: Reinforcement Unlearning for 3D Molecular Flow Matching Models
Lianghong Chen ⋅ Ganlin Feng ⋅ Pingzhao Hu
Self-Improvement with Best-of-N and Retraining: Optimal Scaling for Multiclass SGD
Alireza Mousavi-Hosseini ⋅ Murat Erdogdu ⋅ Adel Javanmard
Self-Improvement with Best-of-N and Retraining: Optimal Scaling for Multiclass SGD
Alireza Mousavi-Hosseini ⋅ Murat Erdogdu ⋅ Adel Javanmard
A Memorization–Generalization Phase Transition for Diffusion Scores
Adrien Schertzer ⋅ Loucas Pillaud-Vivien ⋅ Jean-Christophe Mourrat
A Memorization–Generalization Phase Transition for Diffusion Scores
Adrien Schertzer ⋅ Loucas Pillaud-Vivien ⋅ Jean-Christophe Mourrat
What can linear attention learn from nonlinear teachers in-context?
Mary Letey ⋅ Arman Rysmakhanov ⋅ Yue Lu ⋅ Cengiz Pehlevan ⋅ Jacob Zavatone-Veth
What can linear attention learn from nonlinear teachers in-context?
Mary Letey ⋅ Arman Rysmakhanov ⋅ Yue Lu ⋅ Cengiz Pehlevan ⋅ Jacob Zavatone-Veth
Physics-Informed Diffusion Surrogate Modeling of Stochastic Hodgkin–Huxley Dynamics
Sharvin Goyal ⋅ Qianxun Ren ⋅ Sanjar Salomov ⋅ Laksh Patel
Physics-Informed Diffusion Surrogate Modeling of Stochastic Hodgkin–Huxley Dynamics
Sharvin Goyal ⋅ Qianxun Ren ⋅ Sanjar Salomov ⋅ Laksh Patel
Verbalizable Representations Emerge before Workspace Functions in Language Models
Arjun Dabir ⋅ Ailun Shen ⋅ Michael Chan ⋅ Bangzishu Huang ⋅ Shanduojiao Jiang ⋅ Mohammed Mahfoud
Verbalizable Representations Emerge before Workspace Functions in Language Models
Arjun Dabir ⋅ Ailun Shen ⋅ Michael Chan ⋅ Bangzishu Huang ⋅ Shanduojiao Jiang ⋅ Mohammed Mahfoud
The role of injected noise in diffusion model generalization
Felix Günther ⋅ Vincent Emmerling ⋅ David Kappel
The role of injected noise in diffusion model generalization
Felix Günther ⋅ Vincent Emmerling ⋅ David Kappel
Unsupervised Recursive Reasoners for Combinatorial Problems
Matthias Plasser ⋅ Günter Klambauer ⋅ Sebastian Lehner
Unsupervised Recursive Reasoners for Combinatorial Problems
Matthias Plasser ⋅ Günter Klambauer ⋅ Sebastian Lehner
Towards a Physical Theory of Cross-Lingual Representation Alignment: The RG Hypothesis
Santiago Acevedo ⋅ Giulio Biroli ⋅ Alessandro Laio ⋅ Marco Baroni
Towards a Physical Theory of Cross-Lingual Representation Alignment: The RG Hypothesis
Santiago Acevedo ⋅ Giulio Biroli ⋅ Alessandro Laio ⋅ Marco Baroni
Distilling LLM Feedback for Lean Theorem Proving
Gaëtan Narozniak ⋅ Gérard Biau ⋅ Remi Munos ⋅ Ahmad Rammal ⋅ Pierre Marion
Distilling LLM Feedback for Lean Theorem Proving
Gaëtan Narozniak ⋅ Gérard Biau ⋅ Remi Munos ⋅ Ahmad Rammal ⋅ Pierre Marion
Gauss-Newton drifting extends natural gradient descent
Théo Dumont ⋅ Théo Lacombe ⋅ François-Xavier Vialard
Gauss-Newton drifting extends natural gradient descent
Théo Dumont ⋅ Théo Lacombe ⋅ François-Xavier Vialard
Latent cycle phase as an inductive bias for conditional diffusion
Michele A Copetti ⋅ Samuel Ruiperez-Campillo ⋅ Jorge da Silva Gonçalves ⋅ Sonia Laguna ⋅ Julia Vogt
Latent cycle phase as an inductive bias for conditional diffusion
Michele A Copetti ⋅ Samuel Ruiperez-Campillo ⋅ Jorge da Silva Gonçalves ⋅ Sonia Laguna ⋅ Julia Vogt
Optimal Transport on Graphs for Fine-Grained Text-to-Image Evaluation
Enzo Pinchon ⋅ Ye Zhu ⋅ Johannes Lutzeyer
Optimal Transport on Graphs for Fine-Grained Text-to-Image Evaluation
Enzo Pinchon ⋅ Ye Zhu ⋅ Johannes Lutzeyer
Does Attention-Guided Masking Really Help Object Discovery in Object-Centric Learning?
Youliang Tao ⋅ Yanhua Han ⋅ Bin Zhao ⋅ Juho Kannala ⋅ Joni Pajarinen ⋅ Rongzhen Zhao
Does Attention-Guided Masking Really Help Object Discovery in Object-Centric Learning?
Youliang Tao ⋅ Yanhua Han ⋅ Bin Zhao ⋅ Juho Kannala ⋅ Joni Pajarinen ⋅ Rongzhen Zhao
Reward-Proportional Policy Optimization in Large Combinatorial Spaces
Armaan Sandhu ⋅ Abhijeet Sinha ⋅ Dianbo Liu
Reward-Proportional Policy Optimization in Large Combinatorial Spaces
Armaan Sandhu ⋅ Abhijeet Sinha ⋅ Dianbo Liu
Memorisation and generalisation in auto-regressive generative models
Eleonora Bergamin ⋅ Enrico Ventura ⋅ Antoine Maillard ⋅ Sebastian Goldt
Memorisation and generalisation in auto-regressive generative models
Eleonora Bergamin ⋅ Enrico Ventura ⋅ Antoine Maillard ⋅ Sebastian Goldt
Improved sampling schedules for Discrete Diffusion Models
Alberto Foresti ⋅ Mustapha BOUNOUA ⋅ Giulio Franzese ⋅ Luca Ambrogioni ⋅ Pietro Michiardi
Improved sampling schedules for Discrete Diffusion Models
Alberto Foresti ⋅ Mustapha BOUNOUA ⋅ Giulio Franzese ⋅ Luca Ambrogioni ⋅ Pietro Michiardi
The Geometry of Language Model Logits: Generic Structure and Learned Departures
Swathi Shree Narashiman ⋅ Aravind Ramana Venkatramanan ⋅ Nambirajan Seshadri
The Geometry of Language Model Logits: Generic Structure and Learned Departures
Swathi Shree Narashiman ⋅ Aravind Ramana Venkatramanan ⋅ Nambirajan Seshadri
Surfing the Activation Wave: An Information-Theoretic and Geometric Analysis of LLM Hidden States Applied to Speculative Decoding
Raghavv Goel ⋅ Sudhanshu Agrawal ⋅ Risheek Garrepalli ⋅ Mukul Gagrani ⋅ Mingu Lee
Surfing the Activation Wave: An Information-Theoretic and Geometric Analysis of LLM Hidden States Applied to Speculative Decoding
Raghavv Goel ⋅ Sudhanshu Agrawal ⋅ Risheek Garrepalli ⋅ Mukul Gagrani ⋅ Mingu Lee
Grammar Sparsity Sets How Transformers Learn Random Languages
Edgar Andrés Hernández Moreno ⋅ Francesco Cagnetta
Grammar Sparsity Sets How Transformers Learn Random Languages
Edgar Andrés Hernández Moreno ⋅ Francesco Cagnetta
How Should We Train Tractable Proxies for Language Models?
Hanzhang Liu ⋅ William Zhao ⋅ Zilei Shao ⋅ Benjie Wang ⋅ Guy Van den Broeck
How Should We Train Tractable Proxies for Language Models?
Hanzhang Liu ⋅ William Zhao ⋅ Zilei Shao ⋅ Benjie Wang ⋅ Guy Van den Broeck
Length Generalization Needs Proper Regularization
Pavlo Vasylenko ⋅ André Martins ⋅ Matthias Lindemann ⋅ Marcos Treviso
Length Generalization Needs Proper Regularization
Pavlo Vasylenko ⋅ André Martins ⋅ Matthias Lindemann ⋅ Marcos Treviso