MLRC 2026

Welcome to the home of ML Reproducibility Challenge. This is an annual event promoting research into reproducibility of Machine Learning literature. (v1, v2, v3, v4, v5, v6, v7, v8). This conference is a unique venue in the Machine Learning community to share, disseminate and discuss reproducible methods and tools, investigate reproducibility of papers accepted for publication at top conferences, and test generalizability of scientific findings by adding novel insights and empirical results.

Accepted Papers

We are delighted to announce the 16 papers accepted to the Machine Learning Reproducibility Challenge (MLRC) 2026, an official track of NeurIPS 2026. The papers will be presented in person at NeurIPS 2026 in Sydney, Australia (December 6–13, 2026), alongside papers from the Main Track and the Evaluations & Datasets Track. Each paper was first accepted at the Transactions on Machine Learning Research (TMLR), and subsequently underwent a light compatibility review by the MLRC committee to confirm its suitability for the track.

This announcement was also published on the NeurIPS blog.

The MLRC poster session will be held on Thursday, December 10, 2026, at 5:00-8:00 PM Sydney time. We invite all attendees to visit the poster presentations, meet the authors, and engage in discussion with them.

The Best Paper Award goes to “Evaluating the evidence trace of NeurIPS 2025 contributions with agentic code auditing”. Outstanding Paper Awards go to “Min-p, Max Exaggeration: A Critical Analysis of Min-p Sampling in Language Models” and “A Reproduction Study of Weight-Based Mechanistic Interpretability in Bilinear MLPs”. The award winners are also marked in the list below.

Accepted Papers from the Transactions on Machine Learning Research (TMLR)

1. A Reproduction Study of Weight-Based Mechanistic Interpretability in Bilinear MLPs Itay Erlich, May Ben Zion, Jacob Shashoua (Outstanding Paper Award)
2. Auditing Closed-Loop Learning in Recurrent Neural Networks: Reproduction, Robustness, and Generalization Aarav Sinha
3. [Re] Boosting the Visual Interpretability of CLIP via Adversarial Fine-Tuning Anton Nuzhdin, Andrei Gruzitski, Balázs Egyed, Aleksandr Raudvee
4. Evaluating the evidence trace of NeurIPS 2025 contributions with agentic code auditing Pascal Iversen, Ferdous Nasri, Bernhard Y. Renard, Katharina Baum (Best Paper Award)
5. [Re] FairDICE: A Fair Tradeoff in Multi-objective Offline RL Peter Adema, Karim Galliamov, Aleksey Evstratovskiy, Ross Geurts
6. Min-p, Max Exaggeration: A Critical Analysis of Min-p Sampling in Language Models Rylan Schaeffer, Joshua Kazdan, Yegor Denisov-Blanch (Outstanding Paper Award)
7. Opening Up a New Layer: A Deeper Look into “Interpreting CLIP with Hierarchical Sparse Autoencoders” Nora Kahleogullari, Ada Romanowski, Mathan Sundarrajan, Sally Veerman
8. Reproducibility study of “Bilinear MLPs enable weight-based mechanistic interpretability” Hamid Reza Ahmadi, Mihnea Andrei Bârsan, Jan Jelínek, Robert-Ștefan Sofroni
9. Reproducibility study of FACTER: Fairness-Aware Conformal Thresholding and Prompt Engineering Elena Belkina, Alexandra Elena Antonica, Brigitta Varga, Mattia Sferrazza, Clara Rus
10. Reproducing FACTER: Fairness via Conformal Thresholding and Prompt Repair Oscar Miró López-Feliu, Daimy van Loo, Xanthos Kekkos, Mikel Blom, Clara Rus
11. Revisiting “Edit Away and My Face Will not Stay: Personal Biometric Defense against Malicious Generative Editing” Luis Vitor Zerkowski, Soham Chaudhuri, Finley Helms, Jelle Sombekke, Udit Thakur
12. Revisiting B2T: Discovering and Mitigating Visual Biases through Keyword Explanations Faissal El Kayouhi, Aïda Asma, Joey Laarhoven, Fiona Nagelhout
13. Revisiting RegressionMSR: A Reproducibility Study Neev Mistry
14. Stable Continuous-Time Consistency Distillation: An Empirical Study with a Multistep Extension Vinay Saji Mathew, Soundar Kumara, Gretta Kellogg, William KM Lai
15. WAREX: Web Agent Reliability Evaluation on Existing Benchmarks Su Kara, Fazle Faisal, Suman Nath
16. XFeat Revisited: Reproducibility and Evaluation of a Lightweight Image Matcher Lazar Đoković, Aimee Lin

Venue

We are excited to announce that MLRC 2026 will be an official track at NeurIPS 2026! This marks the first time MLRC is integrated as a dedicated track within NeurIPS. Accepted papers will be presented at NeurIPS 2026 in Sydney, Australia, December 6–13, 2026. As a track within NeurIPS, MLRC is a complementary venue to the Main Track and the Evaluations & Datasets Track.

Important Dates

• Submit to TMLR OpenReview: openreview.net/group?id=TMLR
• Earliest date for consideration: June 21, 2025 00:00 AOE
• Deadline to share your intent to submit a TMLR paper to MLRC: June 4, 2026 23:59 AOE — submit here
• Deadline to have your TMLR decisions in our system: September 30, 2026 AOE
• Announcement of accepted papers / author notifications: October 7, 2026
• NeurIPS 2026: December 6–13, 2026, Sydney, Australia

Organizers

• Koustuv Sinha, Meta
• Jessica Forde, Brown University
• Ana Lucic, University of Amsterdam
• Fernando Pascoal Dos Santos, University of Amsterdam
• Candace Ross, Meta
• Adina Williams, Meta
• Naila Murray, Meta
• Joelle Pineau, McGill University / Mila / Cohere

Contact

• For queries related to the conference, or sponsorship requests, please contact us at reproducibility-chairs@neurips.cc
• Follow us on Social media for updates: Twitter (@repro_challenge), BlueSky (@reproml.org)