❌

Normal view

Received — 27 May 2026 ⏭

EXPO-FT: Sample-Efficient Reinforcement Learning Finetuning for Vision-Language-Action Models

✇cs.AI, q-bio.NC updates on arXiv.org
By: Perry Dong · Kuo-Han Hung · Tian Gao · Dorsa Sadigh · Chelsea Finn
26 May 2026 at 12:00
arXiv:2605.25477v1 Announce Type: cross Abstract: The ability to efficiently and reliably learn new tasks has been a foundational challenge in robotics. Vision-Language-Action (VLA) models have demonstrated strong generalization across diverse manipulation tasks, yet pretrained policies consistently fall short of the reliability required for real-world deployment. Reinforcement learning (RL) fine-tuning offers a promising path to bridge this gap, but existing approaches either train from scratch without fully leveraging pretrained priors, or fine-tune VLAs without achieving the sample efficiency and success rates that practical deployment demands. We present EXPO-FT, a system for stable, sample-efficient RL finetuning of pretrained VLA policies that closes this gap. Our system solves a suite of challenging manipulation tasks, including routing string lights and inserting the plug to light it up, striking a pool ball into a pocket, and inserting a flower into a wine bottle, each requiring combinations of high precision, dynamic actions, and robustness to varied initial states. Our system achieves perfect task performance (30/30 successes) across all evaluated tasks within an average of 19.1 minutes of online robot data, outperforming both prior RL-from-scratch and VLA finetuning approaches. We release an open-source codebase with the aim of facilitating broader adoption of RL finetuning of VLA models in robotics.

Spatial multi-omics defines cancer-associated fibroblasts subtype gradients driving metabolic support and immune remodeling in pancreatic ductal adenocarcinoma

✇(Multiomics OR Omics) AND (Pancreatic)
By: Qiyao Zhang · Zhen Cao · Shangcheng Yan · Bingrong Chen · Bohan Huang · Ziwen Liu · Yupei Zhao
17 May 2026 at 18:00

Cancer Lett. 2026 May 16;653:218585. doi: 10.1016/j.canlet.2026.218585. Online ahead of print.

ABSTRACT

Pancreatic ductal adenocarcinoma is characterized by a fibrotic and metabolically active tumor microenvironment where cancer-associated fibroblasts (CAFs) mediate metabolic crosstalk, extracellular matrix (ECM) remodeling, and immune regulation. However, the metabolic and spatial heterogeneity of CAFs remains incompletely understood. We integrated spatial transcriptomics and spatial metabolomics data from PDAC tissues and performed SpatialGlue-based multimodal clustering to define CAF subtypes. To characterize metabolic communication, we developed an optimal transport (OT)-based metabolic inference framework to quantitatively model metabolite association between CAFs and tumor cells. Subtype-specific features were independently validated using an independent spatial metabolomics cohort and multiplex immunofluorescence (mIHC) staining. Furthermore, these features were correlated with clinical outcomes via TCGA-PAAD deconvolution. Spatial multi-omics integration identified three robust CAF subtypes with distinct signatures. OT analysis revealed differential metabolic interactions: CAF_C0 mediated amino acid/peptide transfer, CAF_C1 was the primary source of lipids, while CAF_C2 exhibited limited metabolic association but stronger immune and ECM signaling activity. Deconvolution confirmed that CAF composition was strongly associated with prognosis; CAF_C2 enrichment predicted poorer survival and gemcitabine resistance, whereas a higher CAF_C0/CAF_C1 balance correlated with improved outcomes. By combining spatial multi-omics with OT-based modeling, this study delineates metabolically and spatially distinct CAF states with clinical relevance. Our findings suggest CAFs act as both metabolic donors and immune-ECM regulators, providing new insights into stromal reprogramming and potential subtype-specific therapeutic targets in PDAC.

PMID:42144098 | DOI:10.1016/j.canlet.2026.218585

Received — 4 April 2026 ⏭

Runtime Burden Allocation for Structured LLM Routing in Agentic Expert Systems: A Full-Factorial Cross-Backend Methodology

✇cs.AI, q-bio.NC updates on arXiv.org
By: Zhou Hanlin · Chan Huah Yong
3 April 2026 at 12:00
arXiv:2604.01235v1 Announce Type: new Abstract: Structured LLM routing is often treated as a prompt-engineering problem. We argue that it is, more fundamentally, a systems-level burden-allocation problem. As large language models (LLMs) become core control components in agentic AI systems, reliable structured routing must balance correctness, latency, and implementation cost under real deployment constraints. We show that this balance is shaped not only by prompts or schemas, but also by how structural work is allocated across the generation stack: whether output structure is emitted directly by the model, compressed during transport, or reconstructed locally after generation. We evaluate this formulation through a comprehensive full-factorial benchmark covering 48 deployment configurations and 15,552 requests across OpenAI, Gemini, and Llama backends. Our central finding is consequential: there is no universal best routing mode. Instead, backend-specific interaction effects dominate performance. Modes that remain highly reliable on Gemini and OpenAI can suffer substantial correctness degradation on Llama, while efficiency gains from compressed realization are strongly backend-dependent. Rather than presenting another isolated model comparison, this work contributes a deployable framework for reasoning about structured routing under heterogeneous backend conditions. We provide a cross-backend evaluation methodology and practical deployment guidance for navigating the correctness-cost-latency frontier in production-grade agentic expert systems.

Do Phone-Use Agents Respect Your Privacy?

✇cs.AI, q-bio.NC updates on arXiv.org
By: Zhengyang Tang · Ke Ji · Xidong Wang · Zihan Ye · Xinyuan Wang · Yiduo Guo · Ziniu Li · Chenxin Li · Jingyuan Hu · Shunian Chen · Tongxu Luo · Jiaxi Bi · Zeyu Qin · Shaobo Wang · Xin Lai · Pengyuan Lyu · Junyi Li · Can Xu · Chengquan Zhang · Han Hu · Ming Yan · Benyou Wang
3 April 2026 at 12:00
arXiv:2604.00986v2 Announce Type: replace-cross Abstract: We study whether phone-use agents respect privacy while completing benign mobile tasks. This question has remained hard to answer because privacy-compliant behavior is not operationalized for phone-use agents, and ordinary apps do not reveal exactly what data agents type into which form entries during execution. To make this question measurable, we introduce MyPhoneBench, a verifiable evaluation framework for privacy behavior in mobile agents. We operationalize privacy-respecting phone use as permissioned access, minimal disclosure, and user-controlled memory through a minimal privacy contract, iMy, and pair it with instrumented mock apps plus rule-based auditing that make unnecessary permission requests, deceptive re-disclosure, and unnecessary form filling observable and reproducible. Across five frontier models on 10 mobile apps and 300 tasks, we find that task success, privacy-compliant task completion, and later-session use of saved preferences are distinct capabilities, and no single model dominates all three. Evaluating success and privacy jointly reshuffles the model ordering relative to either metric alone. The most persistent failure mode across models is simple data minimization: agents still fill optional personal entries that the task does not require. These results show that privacy failures arise from over-helpful execution of benign tasks, and that success-only evaluation overestimates the deployment readiness of current phone-use agents. All code, mock apps, and agent trajectories are publicly available at~ https://github.com/FreedomIntelligence/MyPhoneBench.
Received — 2 April 2026 ⏭

Physiological and Semantic Patterns in Medical Teams Using an Intelligent Tutoring System

✇cs.AI, q-bio.NC updates on arXiv.org
By: Xiaoshan Huang · Conrad Borchers · Jiayi Zhang · Susanne P. Lajoie
1 April 2026 at 12:00
arXiv:2603.29950v1 Announce Type: new Abstract: Effective collaboration requires teams to manage complex cognitive and emotional states through Socially Shared Regulation of Learning (SSRL). Physiological synchrony (i.e., longitudinal alignment in physiological signals) can indicate these states, but is hard to interpret on its own. We investigate the physiological and conversational dynamics of four medical dyads diagnosing a virtual patient case using an intelligent tutoring system. Semantic shifts in dialogue were correlated with transient physiological synchrony peaks. We also coded utterance segments for SSRL and derived cosine similarity using sentence embeddings. The results showed that activating prior knowledge featured significantly lower semantic similarity than simpler task execution. High physiological synchrony was associated with lower semantic similarity, suggesting that such moments involve exploratory and varied language use. Qualitative analysis triangulated these synchrony peaks as ``pivotal moments'': successful teams synchronized during shared discovery, while unsuccessful teams peaked during shared uncertainty. This research advances human-centered AI by demonstrating how biological signals can be fused with dialogues to understand critical moments in problem solving.

Downsides of Smartness Across Edge-Cloud Continuum in Modern Industry

✇cs.AI, q-bio.NC updates on arXiv.org
By: Akhil Gupta Chigullapally · Sharvan Vittala · Razin Farhan Hussian · Mohsen Amini Salehi
1 April 2026 at 12:00
arXiv:2603.29289v1 Announce Type: cross Abstract: The fast pace of modern AI is rapidly transforming traditional industrial systems into vast, intelligent and potentially unmanned autonomous operational environments driven by AI-based solutions. These solutions leverage various forms of machine learning, reinforcement learning, and generative AI. The introduction of such smart capabilities has pushed the envelope in multiple industrial domains, enabling predictive maintenance, optimized performance, and streamlined workflows. These solutions are often deployed across the Industrial Internet of Things (IIoT) and supported by the Edge-Fog-Cloud computing continuum to enable urgent (i.e., real-time or near real-time) decision-making. Despite the current trend of aggressively adopting these smart industrial solutions to increase profit, quality, and efficiency, large-scale integration and deployment also bring serious hazards that if ignored can undermine the benefits of smart industries. These hazards include unforeseen interoperability side-effects and heightened vulnerability to cyber threats, particularly in environments operating with a plethora of heterogeneous IIoT systems. The goal of this study is to shed light on the potential consequences of industrial smartness, with a particular focus on security implications, including vulnerabilities, side effects, and cyber threats. We distinguish software-level downsides stemming from both traditional AI solutions and generative AI from those originating in the infrastructure layer, namely IIoT and the Edge-Cloud continuum. At each level, we investigate potential vulnerabilities, cyber threats, and unintended side effects. As industries continue to become smarter, understanding and addressing these downsides will be crucial to ensure secure and sustainable development of smart industrial systems.

Understanding vs. Generation: Navigating Optimization Dilemma in Multimodal Models

✇cs.AI, q-bio.NC updates on arXiv.org
By: Sen Ye · Mengde Xu · Shuyang Gu · Di He · Liwei Wang · Han Hu
1 April 2026 at 12:00
arXiv:2602.15772v2 Announce Type: replace-cross Abstract: Current research in multimodal models faces a key challenge where enhancing generative capabilities often comes at the expense of understanding, and vice versa. We analyzed this trade-off and identify the primary cause might be the potential conflict between generation and understanding, which creates a competitive dynamic within the model. To address this, we propose the Reason-Reflect-Refine (R3) framework. This innovative algorithm re-frames the single-step generation task into a multi-step process of "generate-understand-regenerate". By explicitly leveraging the model's understanding capability during generation, we successfully mitigate the optimization dilemma, achieved stronger generation results and improved understanding ability which are related to the generation process. This offers valuable insights for designing next-generation unified multimodal models. Code is available at https://github.com/sen-ye/R3.
Received — 26 March 2026 ⏭

UniQueR: Unified Query-based Feedforward 3D Reconstruction

✇cs.AI, q-bio.NC updates on arXiv.org
By: Chensheng Peng · Quentin Herau · Jiezhi Yang · Yichen Xie · Yihan Hu · Wenzhao Zheng · Matthew Strong · Masayoshi Tomizuka · Wei Zhan
25 March 2026 at 12:00
arXiv:2603.22851v1 Announce Type: cross Abstract: We present UniQueR, a unified query-based feedforward framework for efficient and accurate 3D reconstruction from unposed images. Existing feedforward models such as DUSt3R, VGGT, and AnySplat typically predict per-pixel point maps or pixel-aligned Gaussians, which remain fundamentally 2.5D and limited to visible surfaces. In contrast, UniQueR formulates reconstruction as a sparse 3D query inference problem. Our model learns a compact set of 3D anchor points that act as explicit geometric queries, enabling the network to infer scene structure, including geometry in occluded regions--in a single forward pass. Each query encodes spatial and appearance priors directly in global 3D space (instead of per-frame camera space) and spawns a set of 3D Gaussians for differentiable rendering. By leveraging unified query interactions across multi-view features and a decoupled cross-attention design, UniQueR achieves strong geometric expressiveness while substantially reducing memory and computational cost. Experiments on Mip-NeRF 360 and VR-NeRF demonstrate that UniQueR surpasses state-of-the-art feedforward methods in both rendering quality and geometric accuracy, using an order of magnitude fewer primitives than dense alternatives.

Effects of multisensory stimulation based on immersive virtual reality in postoperative neuropsychiatric recovery after gynecological laparoscopy

✇npj Digital Medicine
By: Jiang Liu · Yuxiu Liu · Liuna Bi · Li Zhao · Fuchan Hu · Mengyao Huang · Jingyuan Zhang · Xun Zhou · Ting Wang · Shirong Fang · Fengxian Zhang · Yuanjian Song
24 March 2026 at 08:00

npj Digital Medicine, Published online: 24 March 2026; doi:10.1038/s41746-026-02515-7

Effects of multisensory stimulation based on immersive virtual reality in postoperative neuropsychiatric recovery after gynecological laparoscopy
Received — 17 March 2026 ⏭

LightMoE: Reducing Mixture-of-Experts Redundancy through Expert Replacing

✇cs.AI, q-bio.NC updates on arXiv.org
By: Jiawei Hao · Zhiwei Hao · Jianyuan Guo · Li Shen · Yong Luo · Han Hu · Dan Zeng
16 March 2026 at 12:00
arXiv:2603.12645v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) based Large Language Models (LLMs) have demonstrated impressive performance and computational efficiency. However, their deployment is often constrained by substantial memory demands, primarily due to the need to load numerous expert modules. While existing expert compression techniques like pruning or merging attempt to mitigate this, they often suffer from irreversible knowledge loss or high training overhead. In this paper, we propose a novel expert compression paradigm termed expert replacing, which replaces redundant experts with parameter-efficient modules and recovers their capabilities with low training costs. We find that even a straightforward baseline of this paradigm yields promising performance. Building on this foundation, we introduce LightMoE, a framework that enhances the paradigm by introducing adaptive expert selection, hierarchical expert construction, and an annealed recovery strategy. Experimental results show that LightMoE matches the performance of LoRA fine-tuning at a 30% compression ratio. Even under a more aggressive 50% compression rate, it outperforms existing methods and achieves average performance improvements of 5.6% across five diverse tasks. These findings demonstrate that LightMoE strikes a superior balance among memory efficiency, training efficiency, and model performance.
Received — 11 March 2026 ⏭

Multi-Domain Audio Question Answering Benchmark Toward Acoustic Content Reasoning

✇cs.AI, q-bio.NC updates on arXiv.org
By: Chao-Han Huck Yang · Sreyan Ghosh · Qing Wang · Jaeyeon Kim · Hengyi Hong · Sonal Kumar · Guirui Zhong · Zhifeng Kong · S Sakshi · Vaibhavi Lokegaonkar · Oriol Nieto · Ramani Duraiswami · Dinesh Manocha · Gunhee Kim · Jun Du · Rafael Valle · Bryan Catanzaro
10 March 2026 at 12:00
arXiv:2505.07365v2 Announce Type: replace-cross Abstract: We present Task 5 of the DCASE 2025 Challenge: an Audio Question Answering (AQA) benchmark spanning multiple domains of sound understanding. This task defines three QA subsets (Bioacoustics, Temporal Soundscapes, and Complex QA) to test audio-language models on interactive question-answering over diverse acoustic scenes. We describe the dataset composition (from marine mammal calls to soundscapes and complex real-world clips), the evaluation protocol (top-1 accuracy with answer-shuffling robustness), and baseline systems (Qwen2-Audio-7B, AudioFlamingo 2, Gemini-2-Flash). Preliminary results on the development set are compared, showing strong variation across models and subsets. This challenge aims to advance the audio understanding and reasoning capabilities of audio-language models toward human-level acuity, which are crucial for enabling AI agents to perceive and interact about the world effectively.

Bee: A High-Quality Corpus and Full-Stack Suite to Unlock Advanced Fully Open MLLMs

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yi Zhang · Bolin Ni · Xin-Sheng Chen · Heng-Rui Zhang · Yongming Rao · Houwen Peng · Qinglin Lu · Han Hu · Meng-Hao Guo · Shi-Min Hu
10 March 2026 at 12:00
arXiv:2510.13795v4 Announce Type: replace-cross Abstract: Fully open multimodal large language models (MLLMs) currently lag behind proprietary counterparts, primarily due to a significant gap in data quality for supervised fine-tuning (SFT). Existing open-source datasets are often plagued by widespread noise and a critical deficit in complex reasoning data, such as Chain-of-Thought (CoT), which hinders the development of advanced model capabilities. Addressing these challenges, our work makes three primary contributions. First, we introduce Honey-Data-15M, a new SFT dataset comprising approximately 15 million QA pairs, processed through multiple cleaning techniques and enhanced with a novel dual-level (short and long) CoT enrichment strategy. Second, we introduce HoneyPipe, the data curation pipeline, and its underlying framework DataStudio, providing the community with a transparent and adaptable methodology for data curation that moves beyond static dataset releases. Finally, to validate our dataset and pipeline, we train Bee-8B, an 8B model on Honey-Data-15M. Experiments show that Bee-8B establishes a new state-of-the-art (SOTA) for fully open MLLMs, achieving performance that is competitive with, and in some cases surpasses, recent semi-open models such as InternVL3.5-8B. Our work delivers to the community a suite of foundational resources, including: the Honey-Data-15M corpus; the full-stack suite comprising HoneyPipe and DataStudio; training recipes; an evaluation harness; and the model weights. This effort demonstrates that a principled focus on data quality is a key pathway to developing fully open MLLMs that are highly competitive with their semi-open counterparts.
Received — 5 March 2026 ⏭

WebDevJudge: Evaluating (M)LLMs as Critiques for Web Development Quality

✇cs.AI, q-bio.NC updates on arXiv.org
By: Chunyang Li · Yilun Zheng · Xinting Huang · Tianqing Fang · Jiahao Xu · Lihui Chen · Yangqiu Song · Han Hu
4 March 2026 at 13:00
arXiv:2510.18560v3 Announce Type: replace-cross Abstract: The paradigm of LLM-as-a-judge is emerging as a scalable and efficient alternative to human evaluation, demonstrating strong performance on well-defined tasks. However, its reliability in open-ended tasks with dynamic environments and complex interactions remains unexplored. To bridge the gap, we introduce WebDevJudge, a systematic benchmark for assessing LLM-as-a-judge performance in web development, with support for both non-interactive evaluation based on static observations and continuous interactive evaluation with a dynamic web environment. WebDevJudge comprises human preference labels over paired web implementations, annotated with structured and query-grounded rubrics to ensure high-quality ground truth. Using this benchmark, we comprehensively evaluate various evaluators, including LLMs, MLLMs, and agentic workflows. We systematically investigate the impact of different paradigms and guidance mechanisms. Our experiments reveal a significant gap between LLM judges and human experts. In-depth analysis indicates this gap stems from fundamental model limitations, including failures in recognizing functional equivalence, verifying task feasibility, and mitigating bias. Overall, WebDevJudge presents a challenge to LLM-as-a-judge, offering insights to guide future research toward developing more reliable and capable automated evaluators for complicated scenarios. Code and data are available at https://github.com/lcy2723/WebDevJudge.
There are no more articles
❌