❌

Normal view

Received — 15 September 2026 ⏭

Mr.LHDR: A Benchmark for Multimodal Real-World Long-Horizon Deep Research Agents

✇cs.AI, q-bio.NC updates on arXiv.org
By: Minghao Guo · Meng Cao · Sui Zhao · Siyu Ning · Xin Wang · Haoze Zhao · Jiaxuan Yang · Haihong Hao · Mingfei Han · Shunlin Rong · Haijun Wu · Xiaodan Liang · Xiaojun Chang
14 September 2026 at 12:00
arXiv:2609.11318v2 Announce Type: replace Abstract: Deep research agents are increasingly capable of web search, tool use, multimodal evidence analysis, and information synthesis. However, existing benchmarks mainly evaluate medium-horizon exploration and rarely test whether agents can sustain long, dependency-heavy research processes. We introduce Mr. LHDR (Multimodal real-world Long-Horizon Deep Research), a benchmark for evaluating real-world deep research over long, irreducible chains of interdependent evidence across eight categories. Each question is constructed from a hidden Node-Relation graph and requires an average of 12.1 necessary intermediate conclusions with a mean dependency depth of 10.4 before reaching a short, unique, and verifiable answer. Questions incorporate multimodal evidence, including images, maps, PDFs, logos, charts, tables, and video frames, with at least one non-text element that changes the reasoning state. Mr. LHDR evaluates both final answers and the correctness of intermediate conclusions under annotated dependencies. We evaluate general models, deep research systems, and agent frameworks using Overall Accuracy (OA), Strict Accuracy (SA), Checklist Score (CS), and Dependency-Aware Checklist Score (DACS). Results show that even the strongest system achieves only 43.1% OA and 34.3% SA, indicating that final-answer accuracy substantially overestimates complete research success. Removing images reduces DACS by 12.6 points, demonstrating the importance of multimodal evidence, while SA consistently declines as reasoning chains become longer. These findings reveal sustained, dependency-consistent evidence integration, rather than isolated fact retrieval, as a key bottleneck for current deep research agents.
Received — 10 September 2026 ⏭

Multiomic characterization of malignant pulmonary nodules and development of a methylation-based diagnostic Model

✇Pulmonary nodule
By: Wei Zhou · Tongji Xie · Qingjun Wu · Mengmeng Song · Jianzhen Weng · Jieli Zhang · Xuan Gao · Min Zheng · Zheng Wang · Yunzhi Zhou · Yanming Li
9 June 2026 at 18:00

J Transl Med. 2026 Jun 8;24(1):776. doi: 10.1186/s12967-026-08382-w.

ABSTRACT

BACKGROUND: The molecular distinction between benign and malignant pulmonary nodules remains a significant diagnostic challenge. While genomic drivers are well studied, multiomic integration of the epigenetic-transcriptional landscape and its translation into noninvasive tools are lacking.

METHODS: We performed a multiomic characterization (genomic, epigenomic, and transcriptomic) of 158 pulmonary nodules. Unsupervised factor analysis integrated these layers to identify core regulatory axes. A 9-gene cell-free DNA (cfDNA) methylation classifier was developed and validated in blood and tissue cohorts.

RESULTS: Genomic profiling revealed EGFR mutations (exclusive to malignant nodules) and MYC amplification as fundamental initiators of malignancy. Multiomic factor analysis (Factor 1) revealed profound genetic‒epigenetic synergy, in which these alterations dictate a permissive methylome, leading to aberrant epigenetic programming of chromatin accessibility, as well as epigenetic-transcriptional effects: hypomethylation at the promoters of cell cycle genes that augments their expression, and hypermethylation at immune related pathways gene loci that silences their transcription. This effect orchestrates formation of proproliferative (E2F target/G2M checkpoint) and "immune-cold" malignant phenotype, characterized by elevated Treg/CD8+ ratios and fibroblast recruitment. Notably, we observed a gradual accumulation of methylation aberrations along the premalignant-to-invasive continuum (adenocarcinoma in situ [AIS]→minimally invasive adenocarcinoma [MIA]→adenocarcinoma [ADC]), identifying progressive epigenetic dysregulation as a hallmark of tumor aggressiveness. Global methylome remodeling drives ADC progression through hypermethylation-mediated silencing of tumor suppressors (RASA3 and PPARG) and hypomethylation-activated oncogenic axes, specifically the GDF15 axis, which independently predict poor survival in patients with lung ADC in the TCGA cohort. We translated these tissue-derived insights into a 9-gene cfDNA methylation classifier, which achieved exceptional diagnostic accuracy across independent cohorts (training AUC = 1.00; test AUC = 0.93; tissue AUC = 0.96). Rooted in the biological "ground truth" of tissue dysregulation, this classifier functions specifically as a functional readout of the core cell cycle and proliferative pathways, offering a robust, noninvasive tool for the biology-informed risk assessment of pulmonary nodules.

CONCLUSIONS: This study delineates an epigenetic-transcriptional regulatory network that drives nodule malignancy. Our findings provide a robust theoretical foundation and a high-performance liquid biopsy tool for the precise, noninvasive diagnosis of pulmonary nodules.

PMID:42260586 | PMC:PMC13274191 | DOI:10.1186/s12967-026-08382-w

Received — 27 May 2026 ⏭

Learning from Trials and Errors: Reflective Test-Time Planning for Embodied LLMs

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yining Hong · Huang Huang · Manling Li · Li Fei-Fei · Leonidas Guibas · Jiajun Wu · Yejin Choi
26 May 2026 at 12:00
arXiv:2602.21198v3 Announce Type: replace-cross Abstract: Embodied LLMs endow robots with high-level task reasoning, but they cannot reflect on what went wrong or why, turning deployment into a sequence of independent trials where mistakes repeat rather than accumulate into experience. Drawing upon human reflective practitioners, we introduce Reflective Test-Time Planning, which integrates two modes of reflection: \textit{reflection-in-action}, where the agent uses test-time scaling to generate and score multiple candidate actions using internal reflections before execution; and \textit{reflection-on-action}, which uses test-time training to update both its internal reflection model and its action policy based on external reflections after execution. We also include retrospective reflection, allowing the agent to re-evaluate earlier decisions and perform model updates with hindsight for proper long-horizon credit assignment. Experiments on our newly-designed Long-Horizon Household benchmark and MuJoCo Cupboard Fitting benchmark show significant gains over baseline models, with zero-shot generalization to photorealistic HM3D environments and real-robot experiments on a Franka Panda arm. Ablations confirm that reflection-in-action and reflection-on-action are mutually dependent, and that retrospective reflection achieves better credit assignment than step-wise external feedback at lower computational overhead. Qualitative analyses further highlight behavioral correction through reflection.

ESI-Bench: Towards Embodied Spatial Intelligence that Closes the Perception-Action Loop

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yining Hong · Jiageng Liu · Han Yin · Manling Li · Leonidas Guibas · Li Fei-Fei · Jiajun Wu · Yejin Choi
26 May 2026 at 12:00
arXiv:2605.18746v2 Announce Type: replace-cross Abstract: Spatial intelligence unfolds through a perception-action loop: agents act to acquire observations, and reason about how observations vary as a function of action. Rather than passively processing what is seen, they actively uncover what is unseen - occluded structure, dynamics, containment, and functionality that cannot be resolved from passive sensing alone. We move beyond prior formulations of spatial intelligence that assume oracle observations by recasting the observer as an actor. We introduce ESI-BENCH, a comprehensive benchmark for embodied spatial intelligence spanning 10 task categories and 29 subcategories built on OmniGibson, grounded in Spelke's core knowledge systems. Agents must decide what abilities to deploy - perception, locomotion, and manipulation - and how to sequence them to actively accumulate task-relevant evidence. We conduct extensive experiments on state-of-the-art MLLMs and find that active exploration substantially outperforms passive counterparts, with agents spontaneously discovering emergent spatial strategies without explicit instructions, while random multi-view often adds noise rather than signal despite consuming far more images. Most failures stem not from weak perception but from action blindness: poor action choices lead to poor observations, which in turn drive cascading errors. While explicit 3D grounding stabilizes reasoning on depth-sensitive tasks, imperfect 3D representation proves more harmful than 2D baselines by distorting spatial relations. Human studies further reveal that unlike humans who seek falsifying viewpoints and revise beliefs under contradiction, models commit prematurely with high confidence regardless of evidence quality, exposing a metacognitive gap that neither better perception nor more embodied interaction alone can close.

PRXL2B facilitates the progression of hepatocellular carcinoma and the therapeutic efficacy of oncolytic adenovirus H101 through the PI3K/AKT/PD-L1 axis

✇Omics in Hepatocellular
By: Ziqi Zhang · Hongli Chen · Yangming Liu · Hao Wang · You Yan · Yanxi Yu · Zuotian Huang · Yanyao Liu · Wei Li · Zhongjun Wu · Meiqing Kang · Xiaoyong Luo · Ping Che · Rui Liao
20 May 2026 at 18:00

Biosci Trends. 2026 May 21. doi: 10.5582/bst.2026.01000. Online ahead of print.

ABSTRACT

Oncolytic adenovirus H101 has shown antitumor activity in hepatocellular carcinoma (HCC), but the molecular determinants of treatment response remain unclear. In this study, a Hepa1-6 subcutaneous tumor model was established in C57BL/6 mice and treated with intratumoral H101, followed by integrated transcriptomic and proteomic analyses to identify candidate genes associated with H101 response. PRXL2B was selected for further investigation using public multi-omics datasets, tissue microarray-based immunohistochemistry, in vitro functional assays, mechanistic analyses, and in vivo validation experiments. Integrated multi-omics analyses identified PRXL2B as a candidate gene downregulated after H101 treatment. Public datasets and tissue-based validation further showed that PRXL2B was upregulated in HCC tissues. In MHCC97H and HCCLM3 cells, PRXL2B knockdown inhibited proliferation, migration, and invasion, promoted apoptosis and cell-cycle arrest, and enhanced the antitumor effect of H101. Mechanistically, PRXL2B silencing reduced AKT phosphorylation and PD-L1 expression. In vivo, PRXL2B knockdown suppressed tumor growth, and the combination of PRXL2B knockdown and H101 produced the strongest antitumor effect. These findings indicate that PRXL2B promotes malignant phenotypes in HCC and may modulate H101 efficacy through the PI3K/AKT/PD-L1 axis. Targeting PRXL2B may therefore represent a potential strategy to enhance the therapeutic efficacy of oncolytic virus therapy in HCC.

PMID:42161529 | DOI:10.5582/bst.2026.01000

Received — 13 April 2026 ⏭

Integrated analysis of network pharmacology and multi-omics reveals the mechanisms of Zuogui Jiangtang Qinggan formula ameliorates MASLD via fatty acid metabolic reprogramming

✇(Multiomics OR Omics) AND (Lung OR gastric OR Hepatocellular)
By: Junju Zou · Yuwei Dai · Yungfeng Yu · Yongjun Wu · Qin Xiang · Rong Yu
10 April 2026 at 18:00

Phytomedicine. 2026 Mar 30;155:158128. doi: 10.1016/j.phymed.2026.158128. Online ahead of print.

ABSTRACT

BACKGROUND: The global prevalence of metabolic dysfunction-associated steatotic liver disease (MASLD) continues to rise, and its pathogenesis is complex, creating an urgent need to discover novel and effective therapeutic strategies. The Zuogui Jiangtang Qinggan formula (ZGJTQGF), an approved in-hospital preparation, has demonstrated significant clinical efficacy in treating diabetes over several decades. However, the mechanisms underlying its potential therapeutic effects on MASLD remain unclear PURPOSE: This study systematically investigates the therapeutic effects and molecular mechanisms of ZGJTQGF on MASLD through the integration of network pharmacology and multi-omics strategies.

METHODS: The model of MASLD was successfully induced in db/db mice by a high-fat diet (HFD), which displayed characteristic dyslipidaemia. Serum biomarkers, histology, and hepatic multi-omics analyses were employed to assess metabolic status, steatosis, targets, and pathways. Ultraperformance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS), molecular docking analysis and in vitro verification were applied to explore the active ingredients of ZGJTQGF.

RESULTS: ZGJTQGF significantly reduced dyslipidemia in HFD-fed mice, inhibited pro-inflammatory cytokines, and restored glucose metabolic balance by lowering levels of glucose, insulin, OGTT, and HOMA-IR. Histopathology showed reduced lipid deposition and hepatocyte damage. Comprehensive multi-omics analysis suggested that regulating the AMPK/PGC-1α/PPARα and FXR-BSEP signaling pathways could be potential targets for ZGJTQGF in reprogramming glucose and lipid metabolism in MASLD treatment. Blood component analysis identified 52 ZGJTQGF-derived compounds. In molecular docking experiments, Wogonin, Naringenin, Quercetin, Tanshinone IIA and Berberine showed high-affinity binding to core targets in AMPK, PPARα, PGC-1α, FXR and FAS. Mechanistically, ZGJTQGF activated AMPK/PPARα /PGC-1α and FXR-BSEP signaling pathway, promotes fatty acid β oxidation and enhances energy consumption in AML-2 and 3T3-L1 cells, downregulates SREBP-1-dependent adipogenesis (reduces ACC1 and FAS expression), alleviates MASLD driven reprogramming of glucose and lipid metabolism, and regulates lipid metabolism and fatty acid synthesis.

CONCLUSIONS: ZGJTQGF activates the AMPK/PPARα /PGC-1α pathway and inhibits abnormal lipid accumulation in diabetic fatty liver by promoting fatty acid β-oxidation, energy consumption, and bile acid metabolism. These findings provide new insights into the mechanism of ZGJTQGF in the treatment of diabetic fatty liver disease.

PMID:41962267 | DOI:10.1016/j.phymed.2026.158128

Integrated analysis of network pharmacology and multi-omics reveals the mechanisms of Zuogui Jiangtang Qinggan formula ameliorates MASLD via fatty acid metabolic reprogramming

✇Omics in Hepatocellular
By: Junju Zou · Yuwei Dai · Yungfeng Yu · Yongjun Wu · Qin Xiang · Rong Yu
10 April 2026 at 18:00

Phytomedicine. 2026 Mar 30;155:158128. doi: 10.1016/j.phymed.2026.158128. Online ahead of print.

ABSTRACT

BACKGROUND: The global prevalence of metabolic dysfunction-associated steatotic liver disease (MASLD) continues to rise, and its pathogenesis is complex, creating an urgent need to discover novel and effective therapeutic strategies. The Zuogui Jiangtang Qinggan formula (ZGJTQGF), an approved in-hospital preparation, has demonstrated significant clinical efficacy in treating diabetes over several decades. However, the mechanisms underlying its potential therapeutic effects on MASLD remain unclear PURPOSE: This study systematically investigates the therapeutic effects and molecular mechanisms of ZGJTQGF on MASLD through the integration of network pharmacology and multi-omics strategies.

METHODS: The model of MASLD was successfully induced in db/db mice by a high-fat diet (HFD), which displayed characteristic dyslipidaemia. Serum biomarkers, histology, and hepatic multi-omics analyses were employed to assess metabolic status, steatosis, targets, and pathways. Ultraperformance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS), molecular docking analysis and in vitro verification were applied to explore the active ingredients of ZGJTQGF.

RESULTS: ZGJTQGF significantly reduced dyslipidemia in HFD-fed mice, inhibited pro-inflammatory cytokines, and restored glucose metabolic balance by lowering levels of glucose, insulin, OGTT, and HOMA-IR. Histopathology showed reduced lipid deposition and hepatocyte damage. Comprehensive multi-omics analysis suggested that regulating the AMPK/PGC-1α/PPARα and FXR-BSEP signaling pathways could be potential targets for ZGJTQGF in reprogramming glucose and lipid metabolism in MASLD treatment. Blood component analysis identified 52 ZGJTQGF-derived compounds. In molecular docking experiments, Wogonin, Naringenin, Quercetin, Tanshinone IIA and Berberine showed high-affinity binding to core targets in AMPK, PPARα, PGC-1α, FXR and FAS. Mechanistically, ZGJTQGF activated AMPK/PPARα /PGC-1α and FXR-BSEP signaling pathway, promotes fatty acid β oxidation and enhances energy consumption in AML-2 and 3T3-L1 cells, downregulates SREBP-1-dependent adipogenesis (reduces ACC1 and FAS expression), alleviates MASLD driven reprogramming of glucose and lipid metabolism, and regulates lipid metabolism and fatty acid synthesis.

CONCLUSIONS: ZGJTQGF activates the AMPK/PPARα /PGC-1α pathway and inhibits abnormal lipid accumulation in diabetic fatty liver by promoting fatty acid β-oxidation, energy consumption, and bile acid metabolism. These findings provide new insights into the mechanism of ZGJTQGF in the treatment of diabetic fatty liver disease.

PMID:41962267 | DOI:10.1016/j.phymed.2026.158128

Received — 8 April 2026 ⏭

CoopGuard: Stateful Cooperative Agents Safeguarding LLMs Against Evolving Multi-Round Attacks

✇cs.AI, q-bio.NC updates on arXiv.org
By: Siyuan Li · Zehao Liu · Xi Lin · Qinghua Mao · Yuliang Chen · Haoyu Li · Jun Wu · Jianhua Li · Xiu Su
7 April 2026 at 12:00
arXiv:2604.04060v1 Announce Type: cross Abstract: As Large Language Models (LLMs) are increasingly deployed in complex applications, their vulnerability to adversarial attacks raises urgent safety concerns, especially those evolving over multi-round interactions. Existing defenses are largely reactive and struggle to adapt as adversaries refine strategies across rounds. In this work, we propose CoopGuard , a stateful multi-round LLM defense framework based on cooperative agents that maintains and updates an internal defense state to counter evolving attacks. It employs three specialized agents (Deferring Agent, Tempting Agent, and Forensic Agent) for complementary round-level strategies, coordinated by System Agent, which conditions decisions on the evolving defense state (interaction history) and orchestrates agents over time. To evaluate evolving threats, we introduce the EMRA benchmark with 5,200 adversarial samples across 8 attack types, simulating progressively LLM multi-round attacks. Experiments show that CoopGuard reduces attack success rate by 78.9% over state-of-the-art defenses, while improving deceptive rate by 186% and reducing attack efficiency by 167.9%, offering a more comprehensive assessment of multi-round defense. These results demonstrate that CoopGuard provides robust protection for LLMs in multi-round adversarial scenarios.
Received — 2 April 2026 ⏭

IMPASTO: Integrating Model-Based Planning with Learned Dynamics Models for Robotic Oil Painting Reproduction

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yingke Wang · Hao Li · Yifeng Zhu · Hong-Xing Yu · Ken Goldberg · Li Fei-Fei · Jiajun Wu · Yunzhu Li · Ruohan Zhang
1 April 2026 at 12:00
arXiv:2603.29315v1 Announce Type: cross Abstract: Robotic reproduction of oil paintings using soft brushes and pigments requires force-sensitive control of deformable tools, prediction of brushstroke effects, and multi-step stroke planning, often without human step-by-step demonstrations or faithful simulators. Given only a sequence of target oil painting images, can a robot infer and execute the stroke trajectories, forces, and colors needed to reproduce it? We present IMPASTO, a robotic oil-painting system that integrates learned pixel dynamics models with model-based planning. The dynamics models predict canvas updates from image observations and parameterized stroke actions; a receding-horizon model predictive control optimizer then plans trajectories and forces, while a force-sensitive controller executes strokes on a 7-DoF robot arm. IMPASTO integrates low-level force control, learned dynamics models, and high-level closed-loop planning, learns solely from robot self-play, and approximates human artists' single-stroke datasets and multi-stroke artworks, outperforming baselines in reproduction accuracy. Project website: https://impasto-robopainting.github.io/

MindCube: Spatial Mental Modeling from Limited Views

✇cs.AI, q-bio.NC updates on arXiv.org
By: Qineng Wang · Baiqiao Yin · Pingyue Zhang · Jianshu Zhang · Kangrui Wang · Zihan Wang · Jieyu Zhang · Keshigeyan Chandrasegaran · Han Liu · Ranjay Krishna · Saining Xie · Jiajun Wu · Li Fei-Fei · Manling Li
1 April 2026 at 12:00
arXiv:2506.21458v2 Announce Type: replace Abstract: Can Vision-Language Models (VLMs) imagine the full scene from just a few views, like humans do? Humans form spatial mental models naturally, internal representations of unseen space, to reason about layout, perspective, and motion. Our MindCube benchmark with 21,154 questions across 3,268 images exposes this critical gap, where existing VLMs exhibit near-random performance. Using MindCube, we systematically evaluate how well VLMs build robust spatial mental models through representing positions (cognitive mapping), orientations (perspective-taking), and dynamics (mental simulation for "what-if" movements). We then explore three approaches to help approximate spatial mental models in VLMs, focusing on incorporating unseen intermediate views, natural language reasoning chains, and cognitive maps. The significant improvement comes from a synergistic approach, "map-then-reason", that jointly trains the model to first generate a cognitive map and then reason upon it. By training models to reason over these internal maps, we boosted accuracy from 37.8% to 57.8% (+20.0%). Adding reinforcement learning pushed performance even further to 61.3% (+23.5%). Our key insight is that such scaffolding of spatial mental models, actively constructing and utilizing internal structured spatial representations with flexible reasoning processes, significantly improves understanding of unobservable space.

InCoder-32B: Code Foundation Model for Industrial Scenarios

✇cs.AI, q-bio.NC updates on arXiv.org
By: Jian Yang · Wei Zhang · Jiajun Wu · Junhang Cheng · Shawn Guo · Haowen Wang · Weicheng Gu · Yaxin Du · Joseph Li · Fanglin Xu · Yizhi Li · Lin Jing · Yuanbo Wang · Yuhan Gao · Ruihao Gong · Chuan Hao · Ran Tao · Aishan Liu · Tuney Zheng · Ganqu Cui · Zhoujun Li · Mingjie Tang · Chenghua Lin · Wayne Xin Zhao · Xianglong Liu · Ming Zhou · Bryan Dai · Weifeng Lv
1 April 2026 at 12:00
arXiv:2603.16790v3 Announce Type: replace-cross Abstract: Recent code large language models have achieved remarkable progress on general programming tasks. Nevertheless, their performance degrades significantly in industrial scenarios that require reasoning about hardware semantics, specialized language constructs, and strict resource constraints. To address these challenges, we introduce InCoder-32B (Industrial-Coder-32B), the first 32B-parameter code foundation model unifying code intelligence across chip design, GPU kernel optimization, embedded systems, compiler optimization, and 3D modeling. By adopting an efficient architecture, we train InCoder-32B from scratch with general code pre-training, curated industrial code annealing, mid-training that progressively extends context from 8K to 128K tokens with synthetic industrial reasoning data, and post-training with execution-grounded verification. We conduct extensive evaluation on 14 mainstream general code benchmarks and 9 industrial benchmarks spanning 4 specialized domains. Results show InCoder-32B achieves highly competitive performance on general tasks while establishing strong open-source baselines across industrial domains.
Received — 26 March 2026 ⏭

CaP-X: A Framework for Benchmarking and Improving Coding Agents for Robot Manipulation

✇cs.AI, q-bio.NC updates on arXiv.org
By: Max Fu, Justin Yu, Karim El-Refai, Ethan Kou, Haoru Xue, Huang Huang, Wenli Xiao, Guanzhi Wang, Fei-Fei Li, Guanya Shi, Jiajun Wu, Shankar Sastry, Yuke Zhu, Ken Goldberg, Linxi "Jim" Fan
25 March 2026 at 12:00
arXiv:2603.22435v1 Announce Type: cross Abstract: "Code-as-Policy" considers how executable code can complement data-intensive Vision-Language-Action (VLA) methods, yet their effectiveness as autonomous controllers for embodied manipulation remains underexplored. We present CaP-X, an open-access framework for systematically studying Code-as-Policy agents in robot manipulation. At its core is CaP-Gym, an interactive environment in which agents control robots by synthesizing and executing programs that compose perception and control primitives. Building on this foundation, CaP-Bench evaluates frontier language and vision-language models across varying levels of abstraction, interaction, and perceptual grounding. Across 12 models, CaP-Bench reveals a consistent trend: performance improves with human-crafted abstractions but degrades as these priors are removed, exposing a dependence on designer scaffolding. At the same time, we observe that this gap can be mitigated through scaling agentic test-time computation--through multi-turn interaction, structured execution feedback, visual differencing, automatic skill synthesis, and ensembled reasoning--substantially improves robustness even when agents operate over low-level primitives. These findings allow us to derive CaP-Agent0, a training-free framework that recovers human-level reliability on several manipulation tasks in simulation and on real embodiments. We further introduce CaP-RL, showing reinforcement learning with verifiable rewards improves success rates and transfers from sim2real with minimal gap. Together, CaP-X provides a principled, open-access platform for advancing embodied coding agents.
Received — 11 March 2026 ⏭

RoboPARA: Dual-Arm Robot Planning with Parallel Allocation and Recomposition Across Tasks

✇cs.AI, q-bio.NC updates on arXiv.org
By: Shiying Duan · Pei Ren · Nanxiang Jiang · Zhengping Che · Jian Tang · Zhaoxin Fan · Yifan Sun · Wenjun Wu
10 March 2026 at 12:00
arXiv:2506.06683v4 Announce Type: replace-cross Abstract: Dual-arm robots play a crucial role in improving efficiency and flexibility in complex multitasking scenarios. While existing methods have achieved promising results in task planning, they often fail to fully optimize task parallelism, limiting the potential of dual-arm collaboration. To address this issue, we propose RoboPARA, a novel large language model (LLM)-driven framework for dual-arm task parallelism planning. RoboPARA employs a two-stage process: (1) Dependency Graph-based Planning Candidates Generation, which constructs directed acyclic graphs (DAGs) to model task dependencies and eliminate redundancy, and (2) Graph Re-Traversal-based Dual-Arm Parallel Planning, which optimizes DAG traversal to maximize parallelism while maintaining task coherence. In addition, we introduce the Cross-Scenario Dual-Arm Parallel Task dataset (X-DAPT dataset), the first dataset specifically designed to evaluate dual-arm task parallelism across diverse scenarios and difficulty levels. Extensive experiments demonstrate that RoboPARA significantly outperforms existing planning methods, achieving higher efficiency and reliability, particularly in complex task combinations. Our code is publicly available at https://github.com/AiDuanshiying/RoboPARA.
Received — 5 March 2026 ⏭

Neuro-Symbolic Decoding of Neural Activity

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yanchen Wang · Joy Hsu · Ehsan Adeli · Jiajun Wu
5 March 2026 at 13:00
arXiv:2603.03343v1 Announce Type: new Abstract: We propose NEURONA, a neuro-symbolic framework for fMRI decoding and concept grounding in neural activity. Leveraging image- and video-based fMRI question-answering datasets, NEURONA learns to decode interacting concepts from visual stimuli based on patterns of fMRI responses, integrating symbolic reasoning and compositional execution with fMRI grounding across brain regions. We demonstrate that incorporating structural priors (e.g., compositional predicate-argument dependencies between concepts) into the decoding process significantly improves both decoding accuracy over precise queries, and notably, generalization to unseen queries at test time. With NEURONA, we highlight neuro-symbolic frameworks as promising tools for understanding neural activity.
Received — 25 February 2026 ⏭

On the Equivalence of Random Network Distillation, Deep Ensembles, and Bayesian Inference

✇cs.AI, q-bio.NC updates on arXiv.org
By: Moritz A. Zanger, Yijun Wu, Pascal R. Van der Vaart, Wendelin B\"ohmer, Matthijs T. J. Spaan
24 February 2026 at 13:00
arXiv:2602.19964v1 Announce Type: cross Abstract: Uncertainty quantification is central to safe and efficient deployments of deep learning models, yet many computationally practical methods lack lacking rigorous theoretical motivation. Random network distillation (RND) is a lightweight technique that measures novelty via prediction errors against a fixed random target. While empirically effective, it has remained unclear what uncertainties RND measures and how its estimates relate to other approaches, e.g. Bayesian inference or deep ensembles. This paper establishes these missing theoretical connections by analyzing RND within the neural tangent kernel framework in the limit of infinite network width. Our analysis reveals two central findings in this limit: (1) The uncertainty signal from RND -- its squared self-predictive error -- is equivalent to the predictive variance of a deep ensemble. (2) By constructing a specific RND target function, we show that the RND error distribution can be made to mirror the centered posterior predictive distribution of Bayesian inference with wide neural networks. Based on this equivalence, we moreover devise a posterior sampling algorithm that generates i.i.d. samples from an exact Bayesian posterior predictive distribution using this modified \textit{Bayesian RND} model. Collectively, our findings provide a unified theoretical perspective that places RND within the principled frameworks of deep ensembles and Bayesian inference, and offer new avenues for efficient yet theoretically grounded uncertainty quantification methods.

It Takes a Good Model to Train a Good Model: Generalized Gaussian Priors for Optimized LLMs

✇cs.AI, q-bio.NC updates on arXiv.org
By: Jun Wu · Patrick Huang · Jiangtao Wen · Yuxing Han
24 February 2026 at 13:00
arXiv:2506.00486v4 Announce Type: replace-cross Abstract: Despite rapid progress in large language models (LLMs), the statistical structure of their weights, activations, and gradients-and its implications for initialization, training dynamics, and efficiency-remains largely unexplored. We empirically show that these quantities in LLMs are well modeled by generalized Gaussian (GG) distributions, and introduce a unified, end-to-end optimization framework grounded in this observation. Our contributions are threefold: (1) a GG-based initialization that aligns with trained model statistics, accelerating convergence and improving accuracy; (2) ACT, a progressive activation-constrained training method that reduces redundancy and propagation overhead; and (3) GCT, a gradient-constrained training algorithm that substantially lowers communication cost in distributed training. Experiments across diverse architectures demonstrate consistently smaller, faster models with minimal communication overhead that match or surpass standard baselines. By anchoring LLM optimization in principled statistical modeling, this work advances efficient, scalable, and hardware-aware AI systems.

MoMaGen: Generating Demonstrations under Soft and Hard Constraints for Multi-Step Bimanual Mobile Manipulation

✇cs.AI, q-bio.NC updates on arXiv.org
By: Chengshu Li · Mengdi Xu · Arpit Bahety · Hang Yin · Yunfan Jiang · Huang Huang · Josiah Wong · Sujay Garlanka · Cem Gokmen · Ruohan Zhang · Weiyu Liu · Jiajun Wu · Roberto Mart\'in-Mart\'in · Li Fei-Fei
24 February 2026 at 13:00
arXiv:2510.18316v2 Announce Type: replace-cross Abstract: Imitation learning from large-scale, diverse human demonstrations has been shown to be effective for training robots, but collecting such data is costly and time-consuming. This challenge intensifies for multi-step bimanual mobile manipulation, where humans must teleoperate both the mobile base and two high-DoF arms. Prior X-Gen works have developed automated data generation frameworks for static (bimanual) manipulation tasks, augmenting a few human demos in simulation with novel scene configurations to synthesize large-scale datasets. However, prior works fall short for bimanual mobile manipulation tasks for two major reasons: 1) a mobile base introduces the problem of how to place the robot base to enable downstream manipulation (reachability) and 2) an active camera introduces the problem of how to position the camera to generate data for a visuomotor policy (visibility). To address these challenges, MoMaGen formulates data generation as a constrained optimization problem that satisfies hard constraints (e.g., reachability) while balancing soft constraints (e.g., visibility while navigation). This formulation generalizes across most existing automated data generation approaches and offers a principled foundation for developing future methods. We evaluate on four multi-step bimanual mobile manipulation tasks and find that MoMaGen enables the generation of much more diverse datasets than previous methods. As a result of the dataset diversity, we also show that the data generated by MoMaGen can be used to train successful imitation learning policies using a single source demo. Furthermore, the trained policy can be fine-tuned with a very small amount of real-world data (40 demos) to be succesfully deployed on real robotic hardware. More details are on our project page: momagen.github.io.

AI-driven Large-scale Electron Microscopy enables Whole-tissue Subcellular Digitization

✇cs.AI, q-bio.NC updates on arXiv.org
By: Li Xiao · Liqing Liu · Hongjun Wu · Jiayi Zhong · Xixia Li · Yan Zhang · Junjie Hu · Sun Fei · Ge Yang · Tao Xu
24 February 2026 at 13:00
arXiv:2511.02860v2 Announce Type: replace-cross Abstract: The distribution and interactions of cellular organelles play a critical role in mediating cellular physiology and pathology. Large-scale electron microscopy enables visualization of organelle distribution and interactions at the tissue level with nanometer resolution, but robust and efficient computational analysis tools are lacking. Here, we present a deep learning tool for universal large-scale 2D/3D electron microscopy analysis, DeepOrganelle. This new tool enables high-throughput, cell-resolved spatiotemporal mapping and digitization of organelle distribution and interactions. When applied to spermatogenesis across 12 stages and 22 differentiation status of the germ cells, DeepOrganelle uncovered previously unrecognized, stage-dependent dynamics of mitochondria-endoplasmic reticulum contact sites within one subphase of prophase I during meiosis. It also revealed coordinated organelle redistribution in Sertoli cells towards the blood-testis barrier, digitizing the remodeling dynamics of the tissue. This study demonstrates that DeepOrganelle provides a powerful framework that captures subcellular dynamics at the whole-tissue level.
There are no more articles
❌