The paper introduces WM‑RMoE, a World Model‑based Risk‑aware Mixture‑of‑Experts framework for autonomous overtaking. It uses a learned latent dynamics model to perform multi‑step rollouts, evaluating cumulative risk at the trajectory level, and employs a hierarchical gating mechanism to coordinate long‑, short‑horizon, and rule‑based safety experts. A Gaussian Mixture Model preserves multimodal maneuver branches, improving robustness and preventing behavioral averaging, leading to better safety compliance, decision stability, and generalization in experiments.
By Yongzhi Liu, Sunan Zhang, Jinchang Xu, Jiawei Wang, Yushu Qiu, Chen Lv, Weichao Zhuang
ADGNet introduces an Asymmetric Dual-text Guided Network for infrared small target detection, addressing challenges of pixel-level methods and multimodal approaches that lack regional guidance. It employs an Asymmetric Dual-text Prompt (ADP) with an abstract target prompt and a detailed background prompt, and an Asymmetric Dual-Branch Interaction (ADBI) module to guide visual features separately, followed by an Adaptive Feature Aggregation (AFA) module for dynamic fusion. The authors also create an Asymmetric Image-Text Infrared (AITIR) dataset with asymmetric text annotations for three public datasets, and show that ADGNet outperforms 21 state‑of‑the‑art methods.
By Tongtong Wang, Mingzhu Xu, Chenglong Yu, Jing Wang, Xiaohui Lin, Weili Guan
REFACTOR‑VLA is a wake/sleep framework that learns reusable motor‑program abstractions for vision‑language‑action models. In its sleep phase, it clusters motor‑program fragments using a Behavioral‑Equivalence Kernel derived from rollouts of a latent world model, while its wake phase emits typed lambda terms that a library‑conditioned action decoder consumes. Experiments on LIBERO show that adding an auxiliary supervised contrastive loss during world‑model warm‑up significantly improves clustering quality, outperforming the strongest published baseline across all four task suites.
By Riyaaz Shaik, Chandru Venkataraman
arXiv:2510.01030v2 Announce Type: replace
Abstract: The human ability to translate diverse perceptual and linguistic inputs into structured behavior has been thought to rest on learning robust repres...
By Zach Studdiford, Timothy T. Rogers, Kushin Mukherjee, Siddharth Suresh
The study evaluates large language models for assessing suicide risk in Arabic crisis helpline calls, comparing Arabic and English models. Using de‑identified transcripts from Lebanon’s National Lifeline, the researchers fine‑tuned instruction‑tuned LLMs and transformer encoders, achieving a macro‑F1 of 81.19 and ROC‑AUC of 90.61 for high‑risk calls in Arabic, and 85.00/92.59 in English. The results show that high‑risk calls are more distinguishable than at‑risk calls, and translating to English does not degrade performance, indicating potential for operator‑facing tools.
By Linhai Ma, Rita El Hachem, Mahatab El Hajj, Lilian Ghandour, Samah Fodeh
The paper investigates whether audio‑language models capture paralinguistic cues beyond spoken content. Using the Expresso dataset and four open‑source models, the authors trace how speaking style information is encoded in the late layers of the audio encoder but is degraded before reaching the final output. They find that some models are content‑driven while others are acoustic‑driven, revealing a gap between what is encoded and what is utilized in current audio‑language models.
By Bhuvan Koduru, Dareen Safar B Alharthi, Rita Singh, Bhiksha Raj
The study investigates whether attention weights in Vision‑Language Models (VLMs) accurately reflect model reasoning for visual inputs. Using causal perturbation analysis, it identifies three distinct processing modes—Faithful‑Sufficient, Faithful‑Distributed, and Non‑Focal—indicating heterogeneous visual attention faithfulness. The research also shows that human‑annotated ground‑truth regions align with model attention in only about 60% of cases, highlighting a systematic divergence between model visual reliance and human intuition across VQA, document, and chart tasks.
By Xurui Song, Weishi Wang, Zhongqi Yue, Kuluhan Binici, Tao Bai, Hongxin Shao, Daniel Dahlmeier, Jun Luo
The paper introduces X‑SG$^2$S, a feed‑forward framework that embeds 1D to 3D watermarks into 3D Gaussian Splatting (3DGS) scenes without altering the original rendering pipeline. It splits watermark messages into patches, uses a self‑adaptive gate to choose injection points, and an XD injection head to embed multi‑modal messages into sorted 3DGS points. A learnable gate and XD‑extraction heads recover the hidden messages, achieving robust watermarking with minimal interference to scene fidelity.
By Zihang Cheng, Wentao Bao, Huiping Zhuang, Chun Li, Xin Meng, Ziqian Zeng, Cen Chen, Ming Li, F. Richard Yu
IntroConformal introduces a training‑free Conformal Risk Control framework that offers finite‑sample, distribution‑free factuality guarantees for Large Vision‑Language Models. It uses introspective signals—layer‑wise semantic stability and verification probability derived from the model’s own hidden states—to assess claim factuality. Experiments across multiple LVLM architectures show that IntroConformal meets the conformal risk guarantee while reducing abstention and matching or surpassing external verifier baselines in claim‑level discrimination.
By Md. Atabuzzaman, Christian Alexander, Chris Thomas
Conversation Coach is a voice‑first AI system that lets managers rehearse difficult workplace conversations in a realistic spoken format. It tackles low‑latency interaction, adaptive bot personalities that simulate various employee types, and personalized feedback on content and policy compliance. The authors compare an end‑to‑end speech‑to‑speech model with a cascaded approach, finding the former offers lower latency and cost, while the cascaded model provides better reasoning for coaching quality, and they deployed the cascaded architecture to 40,000+ managers over six months.
By Fanyou Wu, Suraj Maharjan, Ainur Yessenalina, Dennis Xu Chen, Rahul Srivastava, Srinivasan H. Sengamedu
The paper investigates a failure mode in multimodal large language models called multimodal contextual sycophancy, where external text can override conflicting image evidence. A diagnostic set of 998 cases independently varies visual evidence, commonsense priors, and external text to probe when this failure occurs. Experiments across six models show that a System‑2 Visual Arbitration (S2VA) approach, which withholds text from the visual witness, significantly improves performance over direct witness reports, with the best information boundary varying by model and context source.
By Yi-Cheng Lai, Hen-Hsen Huang
The paper introduces a top‑down approach for multi‑person 3D reconstruction from multiple views, using a unified, instance‑centric human‑aware 3D space. Observations from different cameras are lifted into this shared space where geometry, appearance, and semantic cues are jointly encoded, and a spatial contrastive learning strategy aligns features of the same person across views while separating different individuals. The method then regresses SMPL parameters from 3D tokens in a feed‑forward manner, achieving robust, accurate, and efficient reconstruction even under severe occlusions.
By Yuanwang Yang, Buzhen Huang, Zongxuan Ren, Jing Huang, Kun Li
arXiv:2609.01200v1 Announce Type: new
Abstract: When the visual encoder and the language decoder of a vision-language model (VLM) run on different compute nodes, the intermediate visual-token embeddi...
By Reza Heidari, Hamed R. Tavakoli, Juho Kannala
arXiv:2609.01059v1 Announce Type: new
Abstract: As Vision-Language Models (VLMs) tackle dynamic 3D spatial reasoning, ego-motion perception becomes essential to resolve monocular scale ambiguity. How...
By Jiayu Ding, Zhuodong Liu, Lei Zhang, Manyu Xiong, Hongbo Jin, Haoran Tang, Hongbo Zhang, Changen Zhu, Wenbo Xing
arXiv:2609.01147v1 Announce Type: cross
Abstract: Text-rich visual inputs require models that can read, retrieve, and compress language directly in pixel space, yet existing pixel-text encoders strug...
By Chaohao Yuan, Ruifeng Yuan, Zhuoxu Huang, Yu Rong, Hong Cheng, Hou Pong Chan, Chenghao Xiao
arXiv:2609.00709v1 Announce Type: cross
Abstract: Large Vision-Language Models produce fluent image descriptions but offer limited semantic control: users cannot reliably specify whether captions sho...
By Jongyeop Hyun, Taeyoung Kim, Hyounghun Kim
arXiv:2505.13180v3 Announce Type: replace
Abstract: Integrating Large Language Models with symbolic planners is a promising direction for obtaining verifiable and grounded plans, with recent works ex...
By Matteo Merler, Nicola Dainese, Minttu Alakuijala, Giovanni Bonetta, Pietro Ferrazzi, Yu Tian, Bernardo Magnini, Pekka Marttinen
arXiv:2609.00551v1 Announce Type: cross
Abstract: Multimodal memory offers a scalable interface for long-video question answering, but existing methods often retrieve captions, frames, transcripts, s...
By Yijun Chen, Yaqi Zheng, Yanya Li, Boyi Xiao, Buqiang Xu, Shuofei Qiao, Jizhan Fang, Xinle Deng, Yunzhi Yao, Xuehai Wang, Liuxin Zhang, Hui Li, Huajun Chen, Shumin Deng
arXiv:2609.00443v1 Announce Type: cross
Abstract: Language models learn about grammatical number primarily from co-occurrence, and show frequency effects as a result---sometimes taken to indicate tha...
By Zach Studdiford, Kanishka Misra
arXiv:2609.00377v1 Announce Type: cross
Abstract: We present FoldingAgent, an agentic framework for inferring explicit parametric folding programs directly from origami demonstration videos. Our fram...
By Maya Moriya, Sigal Raab, Yael Vinker, Tali Dekel