The study benchmarks foundation models for chest X‑ray classification across a domain shift from adult to pediatric patients. Six foundation models and two CNN baselines were evaluated on a harmonized five‑label task using adult data for adaptation and tested on both adult and pediatric sets. Results show all models experience a substantial performance drop—over 10 percentage points—when applied to pediatric data, and deeper adaptation does not markedly improve cross‑population generalization.
By Sara In\'acio, Hugo Proen\c{c}a, Jo\~ao C. Neves
The paper introduces ACT3D, a vision‑language model that learns to think and explore in 3D environments using only single‑image inputs and action‑based interactions. By fine‑tuning with trajectory supervision and then applying reinforcement learning driven by final‑answer correctness, ACT3D develops an exploration policy that adapts to the question and improves accuracy across six benchmarks, outperforming both proprietary and open‑source baselines while using fewer input frames. The study shows that outcome‑based RL can teach VLMs to decide when and how to acquire visual evidence, leading to better reasoning and spatial understanding.
By Yijiang Li, Bingyang Wang, Nuno Vasconcelos
FeaTok introduces feature-loss tokenization for image generation, training discrete tokens by reconstructing pretrained features instead of pixels. This approach improves generative performance across multiple self-supervised models (SigLIP, SigLIP2, DINOv2, DINOv3). In class-conditional generation, FeaTok reduces gFDr6 by up to 29% at 256×256 resolution and outperforms the strongest unguided continuous baseline with 38× less sampling compute, while also achieving higher human-preference reward-model scores and surpassing text-conditional baselines after only 500 fine-tuning updates.
By Zhang Xiaofeng, Samuel Lavoie, Nicolas Beltran-Velez, Jianhao Yuan, Felix Friedrich, Jakob Verbeek, Aaron Courville, Michal Drozdzal, Adriana Romero-Soriano
SceneReVis is a vision‑grounded, self‑reflective framework for 3D indoor scene synthesis that uses an iterative diagnose‑and‑act loop to detect and fix spatial conflicts such as collisions. It introduces SceneChain‑12k, a large dataset of causal construction trajectories generated via a reverse engineering pipeline, and employs a two‑stage training recipe that moves from supervised fine‑tuning to agentic reinforcement learning, turning the model into an active spatial planner. Experiments show that SceneReVis achieves state‑of‑the‑art performance in high‑fidelity generation and goal‑oriented optimization, with better transfer to underrepresented room categories and potential for out‑of‑distribution room types.
By Yang Zhao, Shizhao Sun, Meisheng Zhang, Yingdong Shi, Xubo Yang, Jiang Bian
MetaExp is a framework that trains language‑model agents to learn reusable experience across tasks by using verifiable cross‑task environment scaling. It introduces two designs—Shared‑Tool families that keep black‑box tools constant and Shared‑Workflow families that preserve latent procedures—allowing agents to acquire operational and procedural experience respectively. After fine‑tuning Qwen3.5‑122B‑A10B on thousands of trajectories from 1,000 task families, MetaExp achieves significant performance gains on benchmarks such as SpreadsheetBench, APEX‑Agents, and AutomationBench, with improvements persisting even when domain‑excluded training is applied.
By Renhao Li, Xinjie Shen, Yang Su, Jianhong Tu, Derek F. Wong, Dayiheng Liu, Min Yang
PanoDiff-SR is a two‑stage pipeline that generates full‑size dental panoramic radiographs (PRs) at diagnostic resolution. First, a denoising diffusion implicit model creates a low‑resolution PR seed, which is then upscaled by a fine‑tuned hybrid attention transformer to a high‑resolution image. The method achieves realistic results, with a Fréchet inception distance of 40.5 and a clinician detection accuracy of 68.5%, closer to chance than previous GAN‑based approaches.
By Sanyam Jain, Bruna Neves de Freitas, Andreas Basse-OConnor, Alexandros Iosifidis, Ruben Pauwels
KNOU‑Flood extends the open AIFL recipe for global streamflow forecasting by retaining a single‑layer LSTM and two‑stage transfer learning while adding a longer 424‑day input window, lead‑aware forecast inputs, and GraphCast as an additional forecast source. The model also introduces a latent‑noise ensemble trained with the fair CRPS, achieving a median KGE′ of 0.65 at lead time 1 and 0.48 at lead time 10 across 2,099 test basins, outperforming AIFL and a 1,024‑unit baseline. The probabilistic head, GraphCast channels, longer window, and mixed fine‑tuning collectively reduce lead‑time decay, improve calibration, and enhance detection of multi‑year floods.
By Minjong Cheon
The study presents an unsupervised domain adaptation model that classifies normal versus abnormal retinal development using optical coherence tomography (OCT) scans from multiple devices and centres. Trained on labelled data from one OCT system and unlabelled data from another, the ResNet50‑based Domain Adversarial Neural Network achieved accuracies between 87.75% and 96.06% across six source‑to‑target pairings and maintained performance on an unseen Heidelberg Spectralis dataset. The approach demonstrates that OCT classification can be transferred across devices without requiring labelled target data, potentially enabling more consistent assessment of retinal development and large‑scale analysis of existing imaging datasets.
By Zhanhan Tu, Nikhil Reddy, Helen J. Kuht, Ha-Jun Yoon, Gail D. E. Maconachie, Ravi Purohit, Prateek Pani, Line Kessel, Jinu Han, Jing Jin, Shui-Hua Wang, Karthik Bharadwaj, Viral Sheth, Michael Hisaund, Rebecca J. McLean, George Garratt, Chetan Kantibhai Patel, Kanmin Xue, Callum Hunt, Zisen Zhou, Ian De Silva, Seema Teli, Sonal Shah, Vanessa Rodwell, Aarti Patel, Irene Gottlob, Yu-Dong Zhang, Zedong Zheng, Girish Varma, Mervyn G. Thomas
NeoHorse-1 is a family of agent-native models that pursue recursive self‑improvement through agentic post‑training. The system uses a heterogeneous model pool with intelligent routing to record capability demands, service tiers, and interactions, converting these records into training examples that preserve reasoning, tool calls, and harness context. By structuring supervised fine‑tuning into a three‑stage curriculum and employing routing‑guided on‑policy distillation, NeoHorse-1 closes an evaluation‑selection‑update loop, raising macro‑average performance from 58.94 to 64.87 on 4B models and from 65.60 to 69.04 on 9B models across eleven benchmarks, while also producing a prefill‑only variant that outputs structured decisions and probabilities for agent workflows.
By NeoHorse Team, Guoliang Cao, Guohao Dai, Tianyu Guo, Kai Han, Hailin Hu, Zihan Jiang, Xiang Kuang, Boxun Li, Yulong Li, Zehua Pei, Yuchuan Tian, Jiamin Wang, Yu Wang, Yunhe Wang, Yihong Wu, Haiyang Xu, Shuo Zhang, Hang Zhou, Siyang Cheng, Jiayu Fan, Wei He, Qingrui Jiao, Hongguang Li, Zhiyuan Li, Runke Liu, Xi Liu, Xinchen Liu, Sinno Jialin Pan, Yi Ren, Liuyang Song, Chenyu Wang, Bei Yu, Quanlu Zhang, Xiangyu Zhang, Mengyu Zheng, Yingjie Zong
The paper evaluates large language model (LLM) judges for radiology report assessment across multiple imaging modalities and anatomical regions, comparing four existing metrics (RadFact, GREEN, FineRadScore, CRIMSON) with a new metric called VERT. GREEN emerges as the most robust among current metrics, while VERT improves correlation with radiologist judgments by up to 26% over GREEN. The study also shows that ensembling and lightweight fine‑tuning of the Qwen3‑30B model can yield additional performance gains and significant latency reductions.
By Federica Bologna, Jean-Philippe Corbeil, Matthew Wilkens, Asma Ben Abacha
The paper introduces STAMP, a timestamp‑balanced message‑passing operator designed for domain adaptation in chronologically split graph data. STAMP replaces standard neighborhood aggregation with temporally aligned propagation, enforcing first‑moment invariance across time while remaining scalable and easy to implement. Experiments on several temporal graph benchmarks, including ogbn‑arxiv, ogbn‑mag, and a temporal‑split Pokec dataset, show that STAMP and its variant STAMP‑mag consistently improve node classification accuracy and achieve state‑of‑the‑art performance.
By Joo Young Park, Sejun Park, Hyunwoo Park
DexMimic is a generalist dexterous hand‑object tracking policy trained across diverse human hand‑object interaction trajectories. It uses Value‑Guided Reset Allocation to adapt the initialization distribution, prioritizing difficult phases and selecting reset states that maximize expected tracking returns. The authors present a benchmark for training, fine‑tuning, and generalization, demonstrating improved tracking success, training efficiency, faster fine‑tuning, and potential zero‑shot generalization to unseen trajectories and objects, and validate the approach on a real‑world dexterous manipulation platform.
By Zhexi Luo, Zhixuan Xu, Rui Yang, Zeyu Xiao, Yiwen Hou, Lin Shao
DefoliationCAM (D-CAM) is a solar‑powered edge AI system that estimates grapevine canopy defoliation in real time. It uses YOLOX‑Nano to locate grapevine cordons in 4K images and a lightweight MobileNet classifier to categorize 320 × 320 pixel patches into four damage levels, then aggregates predictions into a Non‑linear Global Defoliation Severity (NGDS) score. The system processes each image in about 1.5 s on an NVIDIA Jetson Orin Nano, and field deployment over five weeks proved autonomous operation, LoRa communication, and remote monitoring are feasible.
By Fuchen Li, Wei-Zhen Liang, Pin-Chu Lai, Xin Qiao, Jingqiu Chen, Haw Yen, Zonglin Yanga, Chi-En Chiang, Rudolph Avalos, Abbigail L. Rath
FADRA is a frequency‑aware diffusion framework designed for video face restoration (VFR). It combines a pre‑trained text‑to‑video diffusion model with lightweight LoRA adapters and a Low‑Quality Pixel‑Alignment Feature Fusion module to adapt the generative prior to VFR. The method introduces a Repeated Residual Adaptation Head that refines residuals step‑wise using low‑quality latent cues, and employs a Frequency‑Aware Loss to supervise multiple spectral bands, improving facial detail recovery and temporal consistency.
By Jin Jiang, Jia Wang, Panwen Hu, Weiran Zhao, Shengcai Liao
TensorDex is a lossless, tensor‑centric storage system designed to reduce the storage footprint of large language models by performing delta compression at the tensor level. It decomposes models into individual tensors, uses compact fingerprints to predict which tensors can be compressed together, and incrementally clusters tensors to select effective reference bases as new models arrive. In experiments on 2,890 Hugging Face models, TensorDex cuts storage usage by 70.5%, outperforms the previous best system by 37%, and achieves compression and decompression speeds of 22.9 GB/s and 28.4 GB/s, respectively.
By Tingfeng Lan, Zirui Wang, Yunjia Zheng, Zhaoyuan Su, Juncheng Yang, Yue Cheng
CRAFT is a continual learning framework that freezes model weights and learns low‑rank interventions on hidden representations to mitigate catastrophic forgetting. Tasks that are similar share an intervention, determined by how they change the model’s output distribution, and a KL penalty keeps new tasks close to existing behavior. The approach also allows task identification at test time without explicit labels and shows improved performance and reduced forgetting across benchmarks compared to LoRA‑based methods.
By Md Anwar Hossen, Fatema Siddika, Juan Pablo Munoz, Tanya Roosta, Ali Jannesari
The paper investigates whether the tendency of code language models to default to Python when no language is specified is driven by a small, identifiable set of internal components. Using a comprehensive pipeline—including ablation, activation patching, greedy circuit extraction, and significance testing—across 55 model-language configurations, the authors find that this preference is large, grows with model size, and persists after instruction tuning. They also show that the apparent localization of this behavior depends heavily on analysis choices such as component budget and probe design, and that interventions can sometimes alter the model’s output.
By Harsh Kumar, Tanmay Shailesh Joshi, Rahul Maity, Anusa Saha
The paper shows that long‑context drafting can be performed with short‑context drafters that are frozen and run beyond their training length, using two inference‑time mechanisms to keep their cost bounded. By bounding the attention cache and re‑basing rotary position indices, the authors demonstrate that the speedup achieved with a short‑context drafter persists up to two million tokens across multiple target‑drafter pairs and corpora. They prove that when a common index shift leaves the drafter’s computation unchanged, the acceptance gap can be tightly bounded, ensuring consistent performance.
By Aref Jafari, Vansh Bhatia, Matt Qin, Subhajit Dutta Chowdhury, Ben Lee, Vikram Appia, Ali Ghodsi, Mehdi Rezagholizadeh
The paper introduces Noise Policy Optimization (NPO), a method for aligning few-step generative models with black-box reward signals. NPO learns an amortized flow policy that transforms initial Gaussian noise into a reward-tilted distribution, allowing the pretrained generator to produce higher-reward outputs without fine-tuning. The approach achieves superior reward-quality trade-offs across image, language, and protein generation tasks while drastically reducing inference cost compared to test-time search methods.
By Woocheol Shin, Taeyoung Yun, Jiwon Sung, Hyeongyu Kang, Jeongmin Park, Jaewoo Lee, Kyuil Sim, Dohyun Kim, Yutong He, Jinkyoo Park
arXiv:2610. 20870v1 Announce Type: cross Abstract: We present a generative surrogate for charged-current \(\nu_\mu\)-nucleus scattering that can be calibrated to measured differential cross sections without event-level experimental samples.
By Jose Luis Bonilla, Krzysztof M. Graczyk, Rwik Dharmapal Banerjee, Beata E. Kowal, Hemant Prasad, Jan T. Sobczyk