Publications – Miles

Ilana Sebag, Jean-Yves Franceschi, Alain Rakotomamonjy, Alexandre Allauzen, Jamal Atif. On the MIA Vulnerability Gap Between Private GANs and Diffusion Models. 2025. ⟨hal-05236329⟩
Yunzhen Feng, Ariel Kwiatkowski, Kunhao Zheng, Julia Kempe, Yaqi Duan. PILAF: Optimal Human Preference Sampling for Reward Modeling. Forty-Second International Conference on Machine Learning. ICML 2025., Jul 2025, Vancoucer, Canada. pp.16744-16776, ⟨10.48550/arXiv.2502.04270⟩. ⟨hal-05429131⟩
Yunhao Tang, Kunhao Zheng, Gabriel Synnaeve, Rémi Munos. Optimizing Language Models for Inference Time Objectives using Reinforcement Learning. Forty-Second International Conference on Machine Learning (ICML 2025), Jul 2025, Vancoucer, Canada. pp.59066-59085, ⟨10.48550/arXiv.2503.19595⟩. ⟨hal-05429136⟩
Jonas Gehring, Kunhao Zheng, Jade Copet, Vegard Mella, Quentin Carbonneaux, et al.. RLEF: Grounding Code LLMs in Execution Feedback with Reinforcement Learning. Forty-Second International Conference on Machine Learning. ICML 2025., Jul 2025, Vancoucer, Canada. pp.19034-19055, ⟨10.48550/arXiv.2410.02089⟩. ⟨hal-05429109⟩
Alexandre Verine, Florian Le Bronnec, Kunhao Zheng, Alexandre Allauzen, Yann Chevaleyre, et al.. Improving Diversity in Language Models: When Temperature Fails, Change the Loss. Forty-second International Conference on Machine Learning, ICML, Jul 2025, Vancouver, Canada. pp.61266-61300. ⟨hal-05570818⟩
Zheng Kunhao, Decugis Juliette, Gehring Jonas, Cohen Taco, Benjamin Negrevergne, et al.. What Makes Large Language Models Reason In (Multi-Turn) Code Generation?. ICLR, May 2025, Singapore, Singapore. ⟨hal-05070997⟩
Léo Dana, Muni Sreenivas Pydi, Yann Chevaleyre. Memorization in Attention-only Transformers. International Conference on Artificial Intelligence and Statistics, May 2025, Mai Khao, Thailand. pp.3133--3141. ⟨hal-05427727⟩
Blaise Delattre, Paul Caillon, Quentin Barthélemy, Erwan Fagnou, Alexandre Allauzen. Bridging the Theoretical Gap in Randomized Smoothing. International Conference on Artificial Intelligence and Statistics, May 2025, Mai Khao, Thailand. pp.3997-4005. ⟨hal-05415915⟩
Lucas Gnecco Heredia, Matteo Sammut, Muni Pydi, Rafaël Pinot, Benjamin Negrevergne, et al.. Unveiling the Role of Randomization in Multiclass Adversarial Classification: Insights from Graph Theory. AIStats, May 2025, Phuket, Thailand. ⟨hal-05071096v2⟩
Ori Yoran, Kunhao Zheng, Fabian Gloeckle, Jonas Gehring, Gabriel Synnaeve, et al.. The KoLMogorov Test: Compression by Code Generation. The Thirteenth International Conference on Learning Representations. ICLR 2025., Apr 2025, Singapore, Singapore. ⟨10.48550/arXiv.2503.13992⟩. ⟨hal-05429074⟩