The paper tackles two main issues in multi-subject video generation—uncontrollable fidelity strength and semantic drift—by studying Diffusion Transformers (DiTs). It discovers that certain attention blocks naturally create an Intrinsic Spatial Grounding Map (ISGM) that accurately locates reference subjects. Leveraging this insight, the authors introduce Dual-phase Intrinsic Attention Leveraging (DIAL), which uses ISGM during low-noise stages to control fidelity strength without retraining and during high-noise stages to generate preference pairs for reinforcement learning, thereby anchoring attention and reducing semantic drift. Experiments on the OpenS2V-Eval benchmark show that DIAL outperforms baseline models, improving identity consistency and enabling controllable fidelity strength.
FreeFlow is a hierarchical transformer for optical flow estimation that eliminates traditional flow-specific inductive biases such as correlation volumes and feature warping. It relies on a single feed-forward encoder–decoder architecture that incorporates window attention, shifted-window attention, and reduced-resolution global attention. This design allows the model to scale with capacity while achieving state‑of‑the‑art accuracy on Sintel, KITTI‑2015, and Spring benchmarks, all while remaining memory efficient at 1080p inference.
The paper presents a pragmatic segmentation pipeline for brain metastases in the BraTS 2026 Task 1, using a 5‑fold nnU-Net ResEnc‑L ensemble trained for 1,000 epochs on 1,296 four‑modality cases. A rule‑based post‑processing cascade tuned for the lesion‑wise Dice similarity coefficient (LW‑DSC) improves performance, achieving LW‑DSC scores of 0.733, 0.751, 0.713, and 0.549 on enhancing tumour, tumour core, whole tumour, and resection cavity, respectively. The authors audit each post‑processing stage with a five‑fold out‑of‑fold analysis, confirm two stages as robust, and provide a mechanistic analysis of LW‑DSC, along with thirteen negative results that challenge common intuitions.
The paper introduces TASCO, a test‑time adaptation method that improves Large Language Model reasoning by optimizing stability‑aware confidence. It keeps the LLM frozen and uses a lightweight task‑level prefix, applying Random Perturbation for distributional stability and Sharpness‑Aware Perturbation for worst‑case sensitivity. Experiments show enhanced reasoning accuracy and token efficiency across various LLMs and benchmarks, while preserving stable confidence without over‑concentrating predictions.
The paper introduces GenQAS, a tensor‑network‑guided reinforcement learning framework that uses a learned local transition model to generate synthetic circuit transitions for prioritized generative replay. By mixing these synthetic transitions with real experience during Double Deep Q‑Network updates, GenQAS addresses sample starvation in quantum architecture search. Experiments on chemical Hamiltonian benchmarks up to 12 qubits and a 15‑qubit Ising model show significant improvements in success probability and circuit compactness, while a noisy 6‑qubit BeH₂ transfer experiment demonstrates a 92.7% reduction in steps to chemical accuracy.
The paper presents Iterative Sequential Transfer (IST), a method for few-shot multiobjective multitask optimization that addresses the bottleneck of aligning elite solution distributions across tasks. IST treats multitask optimization as a sequence of transfer problems, focusing evaluations on one target task per iteration and using a likelihood-informed prioritization to select the task most ready for knowledge integration. Experiments on benchmark and real-world problems demonstrate IST’s effectiveness under tight evaluation budgets.
Tri-DehazeGS tackles the problem of reconstructing clean 3D scenes from hazy multi‑view images by decoupling the scene and the haze medium. It represents the scene with Gaussian primitives while modeling the haze as an independent view‑shared tri‑plane field, and composes hazy observations through a physical scattering model. The method introduces Medium‑Decoupled Transmittance Gradient Compensation (MD‑TGC) to balance gradients in low‑transmittance regions, leading to improved novel‑view reconstruction on real and synthetic haze benchmarks.
The paper introduces LogiMed‑RoB, a benchmark that tests large language models (LLMs) on hierarchical logical consistency in medical risk‑of‑bias assessments, using 860 randomized controlled trials and 14,820 queries based on Cochrane RoB 2.0 expert logic. It evaluates models across four dimensions—Atomic Consistency, Domain Consistency, Aggregation Consistency, and Evidential Faithfulness—revealing a severe Error Compounding Effect where high atomic accuracy does not translate to end‑to‑end consistency. Experiments on ten state‑of‑the‑art LLMs show that even top performers can collapse to 45.13% overall consistency, with some models nearly failing entirely, and that many models struggle to deduce correct outcomes from retrieved evidence.
The paper introduces EgoGenEval, a benchmark that assesses visual generators’ physical consistency during ego‑motion by measuring Camera Motion Grounding and Scene State Preservation across 1,400 cases and 2,360 target views. It shows that current pose‑free models struggle to maintain both camera motion and scene state, and that pairwise supervision in training does not simultaneously improve these aspects. The authors suggest a trajectory‑centric approach that couples self‑conditioned rollouts with explicit pose and visibility supervision to address these limitations.
The paper introduces LAION-Mobile, a 1‑million‑image dataset of modern smartphone photos with EXIF metadata, to test deepfake detectors in a realistic computational‑photography setting. Twelve state‑of‑the‑art detectors were evaluated on a 9,115‑image subset, revealing that none achieved an AUC above 0.624 on AI‑generated content and that many detectors’ false‑alarm rates spike when thresholds are recalibrated for modern smartphone imagery. The study shows that current detectors fail to maintain both high detection performance and low false‑alarm rates on contemporary smartphone photos, highlighting a critical gap in existing evaluation benchmarks.
The paper introduces ReconPlusGen, a method that injects a reconstruction prior into multi‑view 3D generation. It predicts a point cloud in canonical space from multiple input images, then deterministically injects this geometry into a diffusion process via noise inversion and modulates the noise to maintain generative flexibility for completing unseen areas and refining visible geometry. Qualitative results show reconstruction on benchmark and real‑world images, along with an illustration of the reconstruction‑guided noise initialization and modulation.
The paper introduces belief‑shift branching, a method for placing forks in tree‑structured reinforcement learning rollouts by identifying points where a model’s answer belief changes most. Unlike traditional structural or entropy‑based approaches, belief‑shift uses a probe, logit‑lens depth profile, or learned activation direction to locate pivots in the value curve, incurring minimal computational overhead. Experiments across multiple models and benchmarks show that belief‑shift forking consistently outperforms baseline methods, yielding significant gains in mathematics and code tasks.
The paper reports a six‑month, population‑scale measurement of autonomous language‑model trading agents operating in two production fleets: DX Terminal Pro, with 3,505 user‑funded vaults trading real ETH in Base memecoin markets, and the DXAP live alpha fleet, with 500–599 user‑created agents trading Hyperliquid perpetuals. Across roughly 7.5 million single‑model invocations and 231,638 multi‑tool turns, the study finds that operating layer design, risk sliders, and leaderboard boundaries drive behavior more than strategy text; agents are volatility‑blind in sizing, capture little upside, and show no directional edge compared to a retail benchmark. The analysis includes regression discontinuity, permutation nulls, and a 17‑rule methodology canon to validate the findings.
By T. J. Barton, Chris Constantakis, Patti Hauseman, Annie Mous, Alaska Hoffman, Brian Bergeron, Hunter Goodreau
TTGBench is a new benchmark for temporal graph learning that evaluates both structural evolution and semantic drift in text‑attributed graphs. It includes six real‑world, text‑rich datasets with dual volatility and supports multi‑class and multi‑label temporal node classification, addressing gaps left by existing benchmarks. A comprehensive evaluation of 17 state‑of‑the‑art methods shows a clear divide: TGNNs excel at structural prediction but struggle with semantic tracking, while LLM‑based models perform better on semantic tasks but lag in structural prediction.
By Longfei Ma, Zemin Liu, Fei Wu
The paper introduces Evidence-Aligned Entity Verification (EAEV), a method for detecting entity-level hallucinations in retrieval-augmented generation (RAG). EAEV aligns generated entities with retrieved evidence across three dimensions and uses counterfactual stability analysis to maintain robust alignments when evidence changes. Experiments on multiple RAG benchmarks show that EAEV consistently outperforms existing hallucination detection methods and generalizes well.
By Runsong Jia, Zhen Fang, Mengjia Wu, Jie Lu, Yi Zhang
The paper introduces WAPP, a framework for safely learning positive security Web Application Firewall (WAF) policies from live traffic. It combines trust filtering, deterministic rule synthesis, confidence scoring, and validation to protect against poisoned training data. Experiments on controlled applications show that WAPP improves poisoning resilience and blocks confirmed CRS bypasses, though free‑text fields still pose precision challenges.
By Heba Osama, Zeyad Ahmed, Mohamed Amgad, Ahmed Saafan, Jana Elfeky, Mariam Abdelati, Haitham Ghalwash
The paper introduces CodeTD, a novel method that uses topological data analysis of attention maps from code language models to pre‑execution assess code correctness and detect hallucinations. It quantifies prompt‑generation mismatch through topological patterns and is evaluated on multiple benchmarks (HumanEval, MBPP, BigCodeBench, MultiPL‑E) across five programming languages and ten Code LLMs up to 34B parameters. Results show CodeTD outperforms recent baselines and transfers well between coding benchmarks.
By Daria Voronkova, Ilya Trofimov, Anton Dmitriev, Eduard Tulchinskii, Evgeny Burnaev, Serguei Barannikov
FPicker is a topology-guided framework for filament tracing in low‑signal Cryo‑EM images. It combines a center‑endpoint representation with an open‑curve evolution module to model non‑cyclic connectivity, overcoming limitations of pixel‑wise segmenters, box‑based detectors, sequential trackers, and traditional active contours. On simulated benchmarks, FPicker improves mean spatio‑angular precision by over 40% and reduces topological gap rates by more than 60% under extreme noise, and it achieves state‑of‑the‑art performance on real EMPIAR data after fine‑tuning.
By Tingyin Zhao, Mingtao Huang, Yuan Shen
The paper introduces S$^3$F-Net, a dual‑branch network that fuses spatial and spectral representations for medical image classification. It combines a deep spatial CNN with a shallow spectral encoder, SpectraNet, which uses a learnable SpectralFilter layer to process the full Fourier spectrum efficiently. Evaluated on four medical imaging datasets, S$^3$F-Net consistently outperforms spatial‑only baselines, achieving state‑of‑the‑art accuracy on BRISC2025 and surpassing deeper models on the Chest X‑Ray Pneumonia dataset.
By Md. Saiful Bari Siddiqui, Mohammed Imamul Hassan Bhuiyan
WildRelight is the first in-the-wild dataset designed to evaluate single-image relighting models, featuring high-resolution outdoor scenes captured under strictly aligned, temporally varying natural illuminations paired with high-dynamic-range environment maps. The benchmark demonstrates that state-of-the-art models trained on synthetic data suffer severe domain shifts when applied to real-world imagery. Leveraging the dataset’s temporal structure, the authors introduce a physics-guided inference framework combining Diffusion Posterior Sampling with Temporal Sampling-Aware Test-Time Adaptation, enabling synthetic models to self-supervise and align with real-world statistics on-the-fly.
By Lezhong Wang, Mehmet Onurcan Kaya, Siavash Bigdeli, Jeppe Revall Frisvad