iiinova — where your digital self keeps evolving

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Tian Xia, Minghao Liu, Yiqing Liang, Laixi Shi, Jiayun Wang: Ranking-Aware Prompt Optimization for Multimodal Clinical Diagnosis arxiv.org/abs/2609.40361 arxiv.org/pdf/2609.40361 arxiv.org/html/2609.40361

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Junshu Pan, Zhizhang Fu, Shulin Huang, Yiran Ding, Zifan Cheng, Wenqi Shao, Qiaosheng Zhang, Yue Zhang: Semifactual Credit-Augmented Policy Optimization arxiv.org/abs/2609.40360 arxiv.org/pdf/2609.40360 arxiv.org/html/2609.40360

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Dulhan Jayalath, Oiwi Parker Jones: Removing Timing Shortcuts Improves Non-Invasive Brain-to-Text arxiv.org/abs/2609.40359 arxiv.org/pdf/2609.40359 arxiv.org/html/2609.40359

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Razan El Mais, Ali Chehab, Ibrahim Issa, Razane Tajeddine: Is Weight Tying Still Beneficial for Decoder-Only LLMs in Private Settings Under DP-SGD? arxiv.org/abs/2609.40335 arxiv.org/pdf/2609.40335 arxiv.org/html/2609.40335

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Yanbei Chen, Anirudh Goyal, Raghuraman Krishnamoorthi: Scaling Laws for Looped Mixture of Experts arxiv.org/abs/2609.40316 arxiv.org/pdf/2609.40316 arxiv.org/html/2609.40316

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Sachi Shome, William Eiers: Compression Footprints as Security Signals for Model-Poisoning Defense in Federated Learning arxiv.org/abs/2609.40312 arxiv.org/pdf/2609.40312 arxiv.org/html/2609.40312

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

James Hazelden: Disentangling Computation in Multi-Task Neural Networks with the Green's Operator arxiv.org/abs/2609.40292 arxiv.org/pdf/2609.40292 arxiv.org/html/2609.40292

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Zhangyong Liang, Haibin Ling: PMosFM: Preconditioned Manifold Matching for One-Step Physics-Constrained Generation arxiv.org/abs/2609.40287 arxiv.org/pdf/2609.40287 arxiv.org/html/2609.40287

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Pranjal Aggarwal, Lawrence Keunho Jang, Sean Welleck, Daniel Fried, Ruslan Salakhutdinov, Jing Yu Koh: cua-speedrun: Standardized Benchmarking of the Speed of Computer-Use Agents arxiv.org/abs/2609.40284 arxiv.org/pdf/2609.40284 arxiv.org/html/2609.40284

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Tony Chen, et al.: OpenTSLM TeeMoE: A Unified Time-Series Language Model for Forecasting, Contextual Prediction, and Reasoning arxiv.org/abs/2609.40265 arxiv.org/pdf/2609.40265 arxiv.org/html/2609.40265

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Paul Le Van Kiem, Dario Shariatian, Umut Simsekli, Alain Durmus: Distribution Matching Distillation for Continuous Diffusion Language Models arxiv.org/abs/2609.40235 arxiv.org/pdf/2609.40235 arxiv.org/html/2609.40235

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Anmol Kabra, Swathi Saravana Selvam, Albert Gong, Chao Wan, Christian Belardi, Dongyoung Go, Katie Z. Luo, Kilian Q. Weinberger: PhantomEnvironments: Training LLM Agents in Fictional Worlds arxiv.org/abs/2609.40221 arxiv.org/pdf/2609.40221 arxiv.org/html/2609.40221

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Yuchen Xin, Zhihua Zhang: Near-Linear Accuracy Bounds for Moreau--Yosida Unadjusted Langevin Sampling arxiv.org/abs/2609.40193 arxiv.org/pdf/2609.40193 arxiv.org/html/2609.40193

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Sohail (Neel), Sarkar, Shakuntala Baichoo: Cheap to Draw, Expensive to Trust: Certifying Test-Time Scaling Curves arxiv.org/abs/2609.40190 arxiv.org/pdf/2609.40190 arxiv.org/html/2609.40190

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Tomer Alter, Nir Shlezinger, Michael Segal: MANET-GNN: Learned Decentralized Optimization of Power Allocation in Multi-Channel MANETs arxiv.org/abs/2609.40170 arxiv.org/pdf/2609.40170 arxiv.org/html/2609.40170

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Seonvin Cho, Soohyun Choi, Songnam Hong: Role-Adaptive Policy Optimization for Offline Reinforcement Learning arxiv.org/abs/2609.40149 arxiv.org/pdf/2609.40149 arxiv.org/html/2609.40149

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Yichen Wang, Fanghui Liu, Yudong Chen: From Spectra to Joint Schedules in LLM Pre-training: 3+3(+2) Scaling-Law Regimes arxiv.org/abs/2609.40148 arxiv.org/pdf/2609.40148 arxiv.org/html/2609.40148

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Han Zhong, Yinyu Ye: Policy Iteration Is Not Strongly Polynomial for Deterministic Markov Decision Processes: The Price of Algorithmic Anarchy arxiv.org/abs/2609.40147 arxiv.org/pdf/2609.40147 arxiv.org/html/2609.40147

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Joel Shor: From DNA Design to DNA Slimming: Auditable Agentic Discovery of a Deletion-Only Designer arxiv.org/abs/2609.40143 arxiv.org/pdf/2609.40143 arxiv.org/html/2609.40143

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Seungeun Rho, Jeonghwan Kim, Xue Bin Peng, Sehoon Ha: Game-Guided Skill Discovery through Self-Play for Playable Agent Control arxiv.org/abs/2609.40137 arxiv.org/pdf/2609.40137 arxiv.org/html/2609.40137

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Eftychios Protopapadakis, Konstantinos Makantasis, Konstantinos M. Giannoutakis: Prototype-Rule Neurosymbolic Regularization for Rank-Constrained Tensor Neural Networks under Label Scarcity arxiv.org/abs/2609.40131 arxiv.org/pdf/2609.40131 arxiv.org/html/2609.40131

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Massimo Bini, Anders Christensen, Stephan Alaniz, Judah Goldfeder, Ole Winther, Yann LeCun, Ravid Shwartz-Ziv, Zeynep Akata: Learning Functional Subspaces for Neural Network Compression arxiv.org/abs/2609.40127 arxiv.org/pdf/2609.40127 arxiv.org/html/2609.40127

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Elizaveta Iashchinskaia, Egor Gladin: Scalable Cox Regression via Grouped Risk Sets and Sharper LogSumExp Rates arxiv.org/abs/2609.40120 arxiv.org/pdf/2609.40120 arxiv.org/html/2609.40120

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Nie, Xu, Chen, Ren, Hu, Yang, Zhang: Beyond Model Ranking: Regime Diagnosis for Distributional-Statistical Misspecification in Industrial Time-Series Forecasting arxiv.org/abs/2609.40117 arxiv.org/pdf/2609.40117 arxiv.org/html/2609.40117

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Thede, Chen, Winzeck, Bethge, Akata, Schwarz: Replay on Demand: An Emergent Curriculum for Balancing Adaptation and Forgetting in Continued Pretraining arxiv.org/abs/2609.40089 arxiv.org/pdf/2609.40089 arxiv.org/html/2609.40089

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Khoa Nguyen, Dung T. Nguyen, Thong Huynh, Hoang-Hiep Nguyen-Mau, Anh Nguyen, Minh Ngoc Dinh, Juho Kannala: Accelerated Algorithm for Sparse Regularized Partial Optimal Transport arxiv.org/abs/2609.40075 arxiv.org/pdf/2609.40075 arxiv.org/html/2609.40075

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Keegan Harris, Siddharth Prasad, Asher Trockman, Nika Haghtalab, Michael I. Jordan: Inference Auctions arxiv.org/abs/2609.40070 arxiv.org/pdf/2609.40070 arxiv.org/html/2609.40070

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Junyi Zou, Avrova Donz: LARC: Low-Rank Adaptive Residual Connections for Learning in Frozen Models arxiv.org/abs/2609.40063 arxiv.org/pdf/2609.40063 arxiv.org/html/2609.40063

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Rafael Pereira Eufrazio, Eduardo Fernandes Montesuma, Charles Casimiro Cavalcante: Gromov-Wasserstein Distillation for Inductive Multi-View Embedding arxiv.org/abs/2609.40047 arxiv.org/pdf/2609.40047 arxiv.org/html/2609.40047

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Ayoub Ajarra, Debabrota Basu: Efficient Active Auditing of Multi-Group Fairness with Bias Probes arxiv.org/abs/2609.40034 arxiv.org/pdf/2609.40034 arxiv.org/html/2609.40034

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Maksim Bobrin, Maksim Zhdanov, Dmitry Dylov: Fenchel Tilting: Weighted Correction for Efficient Finetuning of Generative Models arxiv.org/abs/2609.40030 arxiv.org/pdf/2609.40030 arxiv.org/html/2609.40030

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Florent Tariolle, Florian Yger: DashVMC: Real-Time Discrete World Model Control in Geometry Dash arxiv.org/abs/2609.40003 arxiv.org/pdf/2609.40003 arxiv.org/html/2609.40003

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Jiaxi Ye: Learning to Explain While Planning: Rule-Aligned Diffusion Planning for Autonomous Driving arxiv.org/abs/2609.39995 arxiv.org/pdf/2609.39995 arxiv.org/html/2609.39995

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Shakhvalieva, Kharchev, Bezrukov, Fedorova, Bocharov, Oseledets, Ternovskii: What Limits Recursive Reasoning Models: Optimization, Architecture and Test-Time Scaling arxiv.org/abs/2609.39967 arxiv.org/pdf/2609.39967 arxiv.org/html/2609.39967

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Jinshi Liu, Jiahao Li, Pan Liu, Yanfeng Li, Rui Qian, Zhao Tong, Yue Sun, Tao Tan: Reliability-Aware Checkpoint Selection for Domain Generalization arxiv.org/abs/2609.39934 arxiv.org/pdf/2609.39934 arxiv.org/html/2609.39934

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Yuyang Wu, Yufeng Du, Hao Peng: RoPE at the End of Its Rope? Theory, Diagnosis, and Mitigation of Long-Context Failures arxiv.org/abs/2609.39929 arxiv.org/pdf/2609.39929 arxiv.org/html/2609.39929

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Qisheng Zhou, Zhen Xiong, Qiaoyu Tan: TRACE: Trajectory Selection for Parallel Scaling of Search Agents arxiv.org/abs/2609.39912 arxiv.org/pdf/2609.39912 arxiv.org/html/2609.39912

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Nafiseh Payani, Soham Das, G. Anthony Wilson, Anahita Khojandi: Patient-Centered Treatment Planning for Chronic Multimorbidity: A Hierarchical Reinforcement Learning Framework for Preference ... arxiv.org/abs/2609.39911 arxiv.org/pdf/2609.39911 arxiv.org/html/2609.39911

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Syed Nazmus Sakib, Abdul Monaf Chowdhury, Nafiul Haque, Shifat E Arman, Md Mehedi Hasan: Do Better Goal Representations Improve Goal-Conditioned Reinforcement Learning? arxiv.org/abs/2609.39901 arxiv.org/pdf/2609.39901 arxiv.org/html/2609.39901

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Jiaju Wu, Yi Hu, Muhan Zhang: Shared Weights, Selected Computations: How Looped Transformers Route What Each Loop Does arxiv.org/abs/2609.39892 arxiv.org/pdf/2609.39892 arxiv.org/html/2609.39892

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Kevin Yu, Tao Guo, Constantinos Antoniou, Panagiotis Angeloudis: Markovian Dynamics Enforcer: Feasibility Preserving Correction on Learned Dynamics Manifolds arxiv.org/abs/2609.39888 arxiv.org/pdf/2609.39888 arxiv.org/html/2609.39888

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Li, Wang, Wang, Li, Shi, Rostamzadeh, Farnadi, Sugiyama, Zhou: Preemptive LLM Unlearning against Forbidden Capability Acquisition via Gradient Sealing arxiv.org/abs/2609.39866 arxiv.org/pdf/2609.39866 arxiv.org/html/2609.39866

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Stipe Frkovi\'c, Metod Jazbec, Christian A. Naesseth: Fork-dLLM: Avoiding the Flexibility Trap in Diffusion Language Models arxiv.org/abs/2609.39859 arxiv.org/pdf/2609.39859 arxiv.org/html/2609.39859

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Luca Becchetti, Matteo Russo, Ruben Skorupinski: Dimension-Free Rank Lifting from Random Hyperplane Arrangements arxiv.org/abs/2609.39855 arxiv.org/pdf/2609.39855 arxiv.org/html/2609.39855

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Mingyue Ma, Zongbo Han, Changqing Zhang, Guangyu Wang: Predicting Multi-View Rashomon Representation: Can We Learn Where Models Disagree? arxiv.org/abs/2609.39848 arxiv.org/pdf/2609.39848 arxiv.org/html/2609.39848

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Hongwei Zhao, Rui Liu, Yansong Liu: Dynamic LoRA-Experts and Prototype-Ensemble Matching for Class-Incremental Learning arxiv.org/abs/2609.39839 arxiv.org/pdf/2609.39839 arxiv.org/html/2609.39839

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Qipei Chen, Wenye Li, Yule Sun, Ke Wei: Fast Regularized Policy Mirror Descent with One-Step TD Updates arxiv.org/abs/2609.39837 arxiv.org/pdf/2609.39837 arxiv.org/html/2609.39837

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Pierre-Louis Lemaire, Luca Schmidt, Wietze Suijker, Alex Hernandez-Garcia, David Rolnick: RainAtlas: A Multi-Continental Dataset for Precipitation Downscaling arxiv.org/abs/2609.39833 arxiv.org/pdf/2609.39833 arxiv.org/html/2609.39833

000

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · 1h

Giacchetta, Bogani, Barbera, De Toni, Caprio, Pugnana, Passerini: Should I stay or should I show? Learning to selectively disclose information arxiv.org/abs/2609.39818 arxiv.org/pdf/2609.39818 arxiv.org/html/2609.39818

000