The paper introduces a Concept-Integrated Transformer (CIT) that uses a pretrained large language model to generate concept abnormality targets with confidence weights, eliminating the need for manual concept annotation. CIT is applied to mobile sensing data from two longitudinal datasets, achieving the highest F1 score on the AFFECT dataset (0.756) and tying for the highest on a PHQ-9 dataset (0.765). The model’s learned concept scores reveal interpretable behavioral and physiological patterns, such as differences in sleep quantity and quality between high and low negative affect groups.
By Yuning Wang, Iman Azimi, Amir M. Rahmani, Pasi Liljeberg
The paper investigates how block‑diffusion language models can use a constant‑size cache to enable efficient parallel decoding. By employing sequence mixers that summarize completed blocks into a reusable state and a block‑causal training objective, the authors pretrain three 3B block‑diffusion denoisers (attention, Mamba, and hybrid) on 300 B tokens. The resulting state‑space cache remains O(1) in memory and latency regardless of context length, yielding significant speed‑up and memory savings compared to traditional attention‑based caches, especially at very long sequences.
By Vaibhav Singh, Pierre-Andr\'e No\"el, Torsten Scholak, Eugene Belilovsky, Oleksiy Ostapenko
The paper proves that stochastic gradient descent with gradient clipping and additive Gaussian noise (SGD‑CN) converges almost surely under smoothness and bounded noise assumptions, given standard decaying step sizes. The analysis extends to momentum variants such as the stochastic heavy ball and Nesterov's accelerated gradient, showing that careful energy constructions yield similar guarantees. These results provide stronger theoretical foundations for understanding the pathwise behaviour of clipped stochastic gradient methods in both convex and nonconvex regimes.
By Amartya Mukherjee, Jun Liu
arXiv:2609. 12067v1 Announce Type: new Abstract: Channel simulation has recently emerged as a useful component in machine learning systems where samples from a prescribed probability distribution are to be compressed.
By Joseph Rowan, Buu Phan, Ashish J. Khisti
GUIDE is a large‑language‑model driven architecture that elicits and infers human user preferences through conversational Bayesian adaptive sampling and symbolic rule‑based learning. It extends adaptive sampling to a wide range of elicitation questions via a flexible type system and initializes domain‑specific preference models using symbolic representations of world knowledge. In simulated investment portfolio optimization, GUIDE outperforms prior methods, LLM‑only baselines, and its own ablated variants by improving cold‑start performance and reducing recommendation regret during early interactions.
By Anagha Tiwari, Alexander G. Gray, Nick Feamster, Brian Jabarian, Alex Imas, Alex Kale
The paper introduces a machine learning pipeline that predicts 2‑hour peak pedestrian volumes at 101 intersections in Portland, Oregon, using built‑environment, land‑use, and street‑network features from open GIS data. Starting from a Negative Binomial GLM, the authors add feature selection, count‑aware gradient boosting, and repeated cross‑validation, ultimately selecting a histogram‑based gradient boosting model with Poisson loss and L1 Lasso feature selection. This model reduces cross‑validated RMSE by 12% and holdout RMSE by 19% compared to the GLM baseline.
By Bahareh Golchin, Banafsheh Rekabdar, Sirisha Kothuri, Joseph Broach
The paper proposes a reinforcement learning (RL) fine‑tuning framework for electronic health record (EHR) foundation models, treating them as generative policies over patient trajectories. By framing clinical prediction tasks such as hospital readmission as event‑conditioned, time‑windowed reasoning problems and designing time‑aware, rollout‑sensitive rewards, the authors show that RL fine‑tuning consistently outperforms pre‑trained backbones and strong baselines. The approach enables smaller models to surpass larger pre‑trained models in data‑limited settings, induces positive transfer across tasks, and produces trajectories with stronger structural and semantic alignment to ground truth, improving downstream utility.
By Yuxin Xiao, Sheng Zhang, Chandan Singh, Tristan Naumann, Hoifung Poon, Jianfeng Gao, Xiaodong Liu
The paper presents a fast machine unlearning method that uses Hessian analysis to identify correlated training data and applies a closed‑form update rule. This approach achieves an 82× speedup over traditional influence‑function unlearning while maintaining or slightly improving model accuracy. Experiments on seven dataset‑architecture pairs, including CIFAR‑100 with ResNet‑50, show strong forgetting performance and low vulnerability to membership inference attacks.
By Ayushi Thakur, Ruchir Gupta, Amit Kumar Jaiswal, Prayag Tiwari
The paper introduces a causal mediation framework to separate direct discrimination from structural inequality in AI-driven credit decisions. Using Pearl’s natural direct and indirect effects, it presents an identification strategy under treatment‑induced confounding and proposes a doubly‑robust estimator with efficiency guarantees. Empirical analysis of 89,465 mortgage applications shows that about 77% of racial denial disparities stem from financial mediators, while the remaining 23% represents a conservative lower bound on direct discrimination.
By Duraimurugan Rajamanickam
The paper introduces Trusted Multi-view learning with Unified Routing (TMUR), a method that separates view-specific evidence extraction from fusion arbitration in multi-view classification. TMUR employs view-private experts, a collaborative expert, and a unified router that assigns sample-level weights based on global context, along with soft load-balancing and diversity regularization to promote balanced and discriminative expert use. Experiments on 14 datasets show that TMUR consistently improves classification accuracy and reliability compared to 15 recent baselines.
By Yilin Zhang, Cai Xu, Haishun Chen, Ziyu Guan, Wei Zhao
SaFeR-Steer is a progressive multi‑turn alignment framework that uses staged synthetic bootstrapping and tutor‑in‑the‑loop GRPO to train a single student model against adaptive, on‑policy attacks. It introduces Trajectory‑Consistent Summative Reward (TCSR) to ensure that low‑quality turns impact the overall trajectory reward. The authors release the STEER dataset and demonstrate that SaFeR‑Steer significantly improves safety and helpfulness on both single‑turn and multi‑turn benchmarks for Qwen2.5‑VL models.
By Haolong Hu, Hanyu Li, Tiancheng He, Huahui Yi, An Zhang, Qiankun Li, Kun Wang, Yang Liu, Zhigang Zeng
The paper presents a rate‑distortion framework for understanding factual hallucination in closed‑book question answering. It shows that even when a fact is observed, limited memory forces it to be stored approximately, leading to errors that can be bounded by a combination of compression distortion and missing coverage. The authors derive a theoretical lower bound on error and validate it with simulations and probes on modern language models.
By Xi Wang, Shijia Xu, Rongfeng Guo
arXiv:2412.09378v4 Announce Type: replace-cross
Abstract: Clinical trials bridge basic research and clinical application, serving as essential steps in drug development. This review examines clinical...
By Tianyang Wang, Ming Liu, Benji Peng, Xinyuan Song, Charles Zhang, Xintian Sun, Qian Niu, Junyu Liu, Silin Chen, Keyu Chen, Ming Li, Pohsun Feng, Ziqian Bi, Yunze Wang, Yichao Zhang, Cheng Fei, Lawrence K. Q. Yan, Ziyuan Qin, Riyang Bao, Zekun Jiang
arXiv:2609.12723v1 Announce Type: new
Abstract: Floorplans arise in many forms, from vector CAD drawings to raster renderings and sensor-derived density maps. This heterogeneity makes it difficult to...
By Xavier Anad\'on, R\'emi Pautrat, Rui Wang
arXiv:2609.13043v1 Announce Type: new
Abstract: Robust medical image segmentation across imaging modalities is challenging because of large differences in appearance and intensity distributions. Mode...
By Ziliang Hong, Hongyi Pan, Halil Ertugrul Aktas, Andrea Bejar, Elif Keles, Frank H. Miller, Michael B. Wallace, Rajesh N. Keswani, Gorkem Durak, Ulas Bagci
arXiv:2609.12771v1 Announce Type: cross
Abstract: Cross-vehicle federated learning enables vehicles to collaboratively improve perception models while keeping locally collected driving data private....
By Hanju Jang (Yonsei University), Gyeongmin Han (Yonsei University), Sungmin Lee (Yonsei University), Kichang Lee (Yonsei University), Chunghan Lee (Toyota Motor Corporation), JeongGil Ko (Yonsei University)
arXiv:2609.12081v1 Announce Type: cross
Abstract: Mobile manipulation requires perceptual evidence at different spatial scales for base motion and arm control, while the two action modalities remain...
By Guangyu Chen, Qiwei Liang, Shaolong Zhu, Tianxing Chen, Zikuan Xiao, Yifan Xie, Lingfeng Zhang, Ping Luo, Renjing Xu, Wenbo Ding
arXiv:2509.24367v2 Announce Type: replace
Abstract: Deepfake generators evolve rapidly, making exhaustive data collection and repeated retraining impractical. Unlike generic multi-task settings, deep...
By Jinhee Park, Guisik Kim, Choongsang Cho, Junseok Kwon
The paper introduces a multifidelity approach for predicting railway-bogie responses, combining low-fidelity multibody simulation data with high-fidelity roller-rig measurements. It employs a time-delay neural network to model simulation trends and a residual-correction network, constrained by a physics-informed dynamic-balance equation, to adjust for discrepancies. The method achieves a coefficient of determination of 0.8197 and low error metrics on a held-out 385 km/h test case.
By Gyeolhee Lee, Moosun Kim, Taewook Kwon, Jaehun Kim, Dongjin Lee
The paper investigates how to reduce computation in neural networks by combining one‑shot magnitude pruning in a static setting with early exit in an adaptive setting. In a simplified single‑neuron model it proves a concentration theorem for pruning and introduces a conditional perceptron whose excess error decreases as a power of the compute gap, with the exponent increasing as partial and full computations align. The authors extend these results to deep networks, showing how pruning distortions accumulate with depth and deriving a compute‑accuracy trade‑off for frozen‑backbone early exit under a Gaussian process framework, with numerical simulations supporting the theoretical scaling laws.
By Erdem Koyuncu