EgoArgus is a new, human‑annotated dataset that tests visual‑language models (VLMs) as situational assistants in five everyday dialogue‑video scenarios. It evaluates how well VLMs understand and decide when to intervene, especially when visual and textual cues are helpful, irrelevant, or conflicting. The study finds that current VLMs still struggle to reliably act as egocentric assistants and that existing modality‑bias mitigation methods offer limited improvement.
By Yu-Chien Tang, Yu-Hsiang Liu, An-Zi Yen
ReliableRAG is a new framework for Retrieval-Augmented Generation that tackles misinformation in multi‑hop question answering. It extracts structured triples from retrieved documents, evaluates each triple’s reliability by combining semantic relevance to the query with credibility, and keeps only the top‑K reliable, non‑redundant triples. Using these refined triples, the system builds robust reasoning chains that filter out deceptive misinformation and produce accurate, trustworthy answers.
By Jinpu Jiang, Xuan Wu, Wenhao Song, Bo Yang, You Zhou, Hongwei Ge, Heow Pueh Lee, Yanchun Liang, Chunguo Wu
The paper introduces an adaptive triggering mechanism for bias correction in large language model (LLM) reasoning. By framing bias intervention as an online change‑point detection problem, the authors update a CUSUM statistic at each step using either a white‑box next‑token probability signal or a black‑box LLM judge signal, and inject corrective prompts only when the accumulated evidence exceeds a calibrated threshold. Experiments on gpt‑4o‑mini and six open‑weight models show that adaptive black‑box triggering restores most of the accuracy lost by fixed‑interval interventions while reducing the number of corrections, whereas the white‑box signal improves ambiguous‑item accuracy but can hurt disambiguated‑item accuracy due to difficulty distinguishing stereotype reliance from correct evidence.
By Nayoung Kim, Mickey Mancenido, Huan Liu
The paper applies sparse autoencoders to a neutrino foundation model trained on IceCube data, uncovering a validated atlas of physical concepts within the model’s internal representation. Causal analysis shows the direction reconstruction head largely ignores this atlas, whereas an uncertainty head trained on the same representation effectively uses quality and brightness features, improving angular resolution from 20.2° to 3.2° at 20% efficiency. These findings demonstrate that mechanistic interpretability can expose latent physics and guide the design of downstream tasks.
By Rapha\"el Bonnet-Guerrini, Johann Ioannou-Nikolaides, Inar Timiryasov, Vincenzo Piuri
The paper investigates how adversarial examples transfer between client models in federated learning and explores the relationship between these examples and client data distributions. It proposes a defense strategy based on adversarial training that leverages the transferability of model robustness. Experiments on real-life datasets demonstrate that the new attack and defense methods outperform existing state‑of‑the‑art approaches.
By Zuobin Xiong, Deval Mukherjee, Homook Cho, Wei Li
The paper surveys Multi‑Modal Anomaly Detection (MMAD), a field that identifies rare abnormal events across heterogeneous data sources used in safety‑critical domains like industrial inspection and cybersecurity. It formalizes MMAD, outlines five core characteristics, and categorizes existing methods into normality‑assumption and anomaly‑assumption paradigms, highlighting how foundation models are reshaping the field. The survey also compiles benchmarks, evaluation protocols, and identifies open problems for developing robust, adaptive, and interpretable MMAD systems.
By Xudong Mou, Zexin Wu, Chuan Luo, Shiru Chen, Xudong Liu, Chunming Hu, Renyu Yang
The paper introduces a hierarchical multimodal mixture-of-experts (MoE) model for interstitial lung disease (ILD) classification. It combines a frozen, pre‑trained imaging expert with structured electronic health records (EHR) through a two‑stage gating system: a modality‑level gate weights imaging and EHR predictions, while a sub‑gating module further decomposes the EHR branch into clinically defined feature groups with learned, group‑specific contributions. The approach preserves stable imaging representations, allows input‑dependent clinical weighting, and enhances interpretability across anatomical regions, imaging–EHR utilization, and EHR feature groups, achieving the highest mean AUC (0.8750 ± 0.0443) under strict patient‑level cross‑validation.
By Alec K. Peltekian, Gorkem Durak, Halil Ertugrul Aktas, Carrie Lynn Richardson, Mary Carns, Kathleen Aren, GR Scott Budinger, Anthony J. Esposito, Alexander Misharin, Alok Nidhi Choudhary, Ankit Agrawal, Ulas Bagci
The paper introduces PuRF, a cache‑based method for multi‑label test‑time adaptation of vision‑language models. PuRF first performs region purification to reliably identify object‑specific regions, then uses these purified regions to build a discriminative, temporally refreshed cache that mitigates dominant‑label bias. Experiments show that PuRF outperforms existing methods, achieving a 4.05% mAP gain on ViT‑B/32 across five datasets.
By Yiwen Liang, Hui Chen, Yizhe Xiong, Mengyao Lyu, Yuhan Cao, Zijia Lin, Shuaicheng Niu, Sicheng Zhao, Jungong Han, Guiguang Ding
VGA‑BenchV2 is an expanded, human‑aligned benchmark and optimization framework that jointly evaluates video generation quality and aesthetic value. It builds on the original VGA‑Bench taxonomy, adding 52 sub‑dimensions and 1,016 curated prompts to generate over 60,000 videos from 12 mainstream models. The benchmark significantly enlarges human supervision with 36,000 task‑level annotations and introduces a hybrid evaluator (VAQA‑Net, VTag‑Net, VGQA‑Net) that aligns well with human judgments and can be used as a reward model for reinforcement‑learning fine‑tuning.
By Longteng Jiang, DanDan Zheng, Qianqian Qiao, Heng Huang, Huaye Wang, Yihang Bo, Bao Peng, Jingdong Chen, Jun Zhou, Xin Jin
SMART is a new framework that jointly tackles continuous sign language recognition (CSLR) and spotting by leveraging a multimodal large language model (MLLM) to generate motion descriptions as auxiliary semantic cues. It performs stable video‑text alignment with small batch sizes and introduces a Multi‑Scale Temporal Adapter to capture temporal interactions during transformer encoding. The framework also incorporates CSFormer, a CSLR‑guided spotting module that injects recognition‑derived gloss evidence into a boundary‑aware spotting network, enabling mutual benefit between recognition and spotting tasks.
By Eunjee Choi, JungHoon Sung, Seongwhan Cho, Chu Xin, Younggeun Choi
MMEmb-R1 is a multimodal embedding framework that enhances reasoning by treating it as a latent variable and selecting beneficial reasoning paths through pair-aware selection and counterfactual intervention. It uses reinforcement learning to invoke reasoning only when necessary, reducing unnecessary computation and latency. On the MMEB-V2 benchmark, MMEmb-R1 achieves a state‑of‑the‑art score of 71.2 with just 4 B parameters.
By Yuchi Wang, Dingkang Yang, Haiyang Yu, Weikang Bian, Jiefeng Long, Xiao Liang, Chao Feng, Hongsheng Li
The paper introduces a term‑resolved framework that decomposes the second moment of the gradient in quantum circuit training into three components: pre‑cancellation activity, sign organization, and their statistical coupling. Using a conditional random‑sign model as a reference, the authors analyze hardware‑efficient and Hamiltonian variational ansätze on transverse‑field and longitudinal‑field Ising models, finding that activity dominates gradient suppression in the hardware‑efficient ansatz while both activity and organization grow comparably in the Hamiltonian variational ansatz. Microscopic sign‑alignment analysis reveals sector‑structured organization in the latter and only weak residual sign structure in the former, offering a detailed characterization of barren plateau behavior beyond aggregate statistics.
By Pilsung Kang
The paper investigates the thermodynamic cost of inference and learning in physical neural networks. It shows that quasi‑static inference requires no work, while finite‑speed inference incurs work bounded by the Wasserstein‑2 distance between thermal states, roughly $k_B T$ per dimension of the widest layer. Learning, however, has an irreducible cost of a few $k_B T$ per parameter, independent of speed, indicating that memory dominates the thermodynamic price.
By Alexei V. Tkachenko
The paper presents a reinforcement learning approach to enhance large language model (LLM) auditors for alignment tasks. By training policies that investigate target models for hidden behaviors and using an LLM judge to compare investigations, the method improves audit realism and reduces false positives. Experiments show better performance on adversarially fine‑tuned targets and a low false‑positive rate below 1%.
By Paul Rosu, Rowan Wang
The paper studies finite‑horizon minimum‑energy steering of inertial swarms under stochastic disturbances, focusing on mean‑field models with Cucker–Smale alignment or Morse attraction–repulsion interactions. It formulates the problem as a Schr"odinger bridge, deriving nonlinear, time‑symmetric optimality systems and proposing nested fixed‑point schemes for numerical solution. Numerical experiments demonstrate that the optimal corrective drift can either exploit or counteract the natural interaction forces, depending on their alignment with the steering objective.
By Asmaa Eldesoukey, Md Zulfiqur Haider, Italo Napolitano, Yongxin Chen, Abhishek Halder
The paper introduces a weak Gromov-Wasserstein (wGW) framework that compares source relations with relations between target conditional laws, focusing on inner-product relations and preserving conditional means. It defines the barycentric weak inner-product GW (wIGW) distance, proves existence of minimizers under finite second moments, and presents a ridge-regularized dual formulation leading to an iterative algorithm for finitely supported measures. Experiments on point clouds, graphs, and a PBMC multiome study demonstrate that mean-preserving target refinements can incur zero cost and improve atlas-based cell type transfer.
By Youssef Mroueh
The paper introduces DeltaGateNet, a framework that models bidirectional temporal dynamics in EEG signals for driving fatigue recognition. It uses a Bidirectional Delta module to separate positive and negative temporal differences, and a Gated Temporal Convolution module to capture long‑term dependencies while preserving channel specificity. Experiments on SEED‑VIG and SADT datasets show that DeltaGateNet outperforms existing methods, achieving high intra‑subject and inter‑subject accuracies across balanced and unbalanced data.
By Yip Tin Po, Jianming Wang, Yutao Miao, Jiayan Zhang, Yunxu Zhao, Xiaomin Ouyang, Zhihong Li, Nevin L. Zhang
The paper introduces RefineCut, an open‑weight planner that edits a typed video timeline by applying structured patches for clip selection, trimming, ordering, transitions, music alignment, and duration. A deterministic verifier checks each patch against an explicit constraint ledger, and the planner is trained via verifier‑replayed distillation and a second evolutionary stage (RefineCut‑Evo) that uses the verifier and a task rubric to generate high‑margin preference pairs. On the RefineCut‑Bench dataset, the 8‑billion‑parameter planner improves from a Video‑Editing Score of 0.620 to 0.924, matching or exceeding its frontier teachers in a closed verifier loop, and the gains transfer to other large models such as Llama‑3.1‑8B and GLM‑4‑9B.
By Haoyu Wang, Cheng Feng, Liuyang Bian, Ruiyang Huang, Lei Wei, Yafei Wen, Xiaoxin Chen, Xiaoying Tang
The paper presents a variational learning framework that simultaneously infers non‑parametric interaction kernels and environmental or intra‑agent forces in collective dynamics. It extends existing methods to handle both interaction and environmental components, validating the approach on benchmark models such as synchronization, alignment, and attraction‑repulsion systems. A model‑selection procedure is also introduced to identify the best explanatory framework from trajectory data, enabling direct recovery of mechanistic interaction mechanisms.
By Nipuni de Silva, Ming Zhong, James M. Greene
RoofDiT is a generative framework that synthesizes and reconstructs 2D roof graphs by modeling roofs as vertex-edge planar graphs. It uses a two-stage diffusion transformer to generate roof vertices and an edge prediction module to infer graph topology, incorporating relative geometry-aware attention, footprint and aerial-image conditioning, and an alignment regularizer. The model supports unconditional generation, footprint-conditioned synthesis, and image-guided reconstruction, achieving superior graph generation quality and edge F1 scores compared to baselines.
By Daniel Panangian, Ksenia Bittner