❌

Normal view

Received — 27 May 2026 ⏭

A Sober Look at Agentic Misalignment in Automated Workflows

✇cs.AI, q-bio.NC updates on arXiv.org
By: Wenqian Ye · Bo Yuan · Zhichao Xu · Yijun Tian · Yawei Wang · Henry Kautz · Aidong Zhang
26 May 2026 at 12:00
arXiv:2605.24197v1 Announce Type: new Abstract: We study a class of emergent misalignment in multi-agent systems (MAS), with a focus on automated workflows, which we refer to agentic misalignment. Although these systems can solve complex tasks, they often fail because agents act according to implicit proxy utilities that do not align with the intended human goals. We formally define these behaviors and analyze them within a Bayesian framework, showing that generic utilities naturally lead to posterior collapse of agents in automated workflows. To address this issue, we propose Agentic Evidence Attribution (AEA), a novel alignment paradigm that improves agent posteriors using context-specific evidence. AEA reasons over agent actions and provides structured evidence to correct misaligned behavior during collaboration. To better understand the role of evidence, we study two instantiations of AEA: self-reflection (internal evidence from the model) and weak-to-strong generalization (external evidence on the agentic trajectory). We show that a small evidence model effectively aligns the MAS by providing orthogonal failure attribution. Our results clarify the sources of agentic misalignment in automated workflows and show that evidence-based alignment can effectively improve agent collaboration and leads to reliable multi-agent systems built on automated workflows.

A Large-Scale Dataset and Benchmark: Do Protein-Ligand Models Learn Binding Sites or Just Binding Likelihood?

✇cs.AI, q-bio.NC updates on arXiv.org
By: Zhaohan Meng · Zhen Bai · Ke Yuan · Iadh Ounis · Zaiqiao Meng · Hao Xu · Joseph Loscalzo
26 May 2026 at 12:00
arXiv:2605.24045v1 Announce Type: cross Abstract: Protein-ligand modeling underpins computational drug discovery and molecular design. Existing protein-ligand benchmarks typically evaluate whether a protein and ligand interact and how strongly they bind, through tasks such as binary binding prediction and affinity regression. However, these evaluations provide limited evidence of whether models can localize binding sites or identify the non-covalent interactions underlying molecular recognition. To address this gap, we introduce InteractBind, a large-scale protein-ligand dataset comprising approximately 100k protein-ligand pairs, together with a benchmark for fine-grained evaluation. The core fine-grained task is that of binding-site localization, which uses protein-residue and ligand-atom interaction maps spanning six major types of non-covalent interactions to assess whether model-derived interaction maps localize binding sites. InteractBind further includes binding affinity and protein similarity-controlled splits to support realistic generalization assessment. Using InteractBind, we evaluate eight existing sequence-based and interaction-aware models, assessing binary binding prediction and binding-site localization. Results reveal limited binding-site localization despite strong binary binding prediction, with marked variation across non-covalent interaction types. Overall, InteractBind establishes a benchmark paradigm that encourages the development of more interpretable and physically grounded protein-ligand models.

FoodMonitor: Benchmarking MLLMs for Explainable Compliance Analysis

✇cs.AI, q-bio.NC updates on arXiv.org
By: Ruihao Xu · Xingming Shui · Jingxuan Niu · Yiqin Wang · Jilin Yu · Haoji Zhang · Yansong Tang
26 May 2026 at 12:00
arXiv:2605.24503v1 Announce Type: cross Abstract: As AI-powered compliance monitoring becomes increasingly important in public governance and industrial safety, the ability to provide verifiable evidence and traceable accountability signals is essential. However, existing video anomaly detection datasets focus on event-level binary classification, lacking the rule-driven, explainable analysis required for real-world compliance scenarios. We introduce FoodMonitor, a benchmark for explainable compliance analysis in commercial kitchen surveillance. FoodMonitor comprises 477 video clips with 3,307 violation annotations across a dual-channel design covering both person-level and environment-level violations. Each annotation specifies which rule was violated, what non-compliant behavior occurred, and who committed it with frame-level bounding boxes. We establish a unified evaluation protocol with a two-stage matching mechanism that separately assesses spatial localization and semantic understanding, along with a composite metric ($C_{\text{score}}$) that balances environment and person detection performance. Systematic evaluation of several state-of-the-art multimodal large language models reveals that the best-performing model achieves only 0.360 $C_{\text{score}}$, with spatial localization and fine-grained rule understanding emerging as the primary bottlenecks. Our analysis identifies two distinct failure modes: localization-dominated errors and semantics-dominated errors, providing diagnostic insights for future model development.

DisDop: Distillation with Domain Priors for Open-Vocabulary Aerial Object Detection

✇cs.AI, q-bio.NC updates on arXiv.org
By: Ruihao Xu · Yong Liu · Yansong Tang · Sule Bai · Xubing Ye · Bingyao Yu · Yutao Guo · Jiwen Lu · Jie Zhou
26 May 2026 at 12:00
arXiv:2605.24639v1 Announce Type: cross Abstract: With the widespread application of drones in recent years, object detection of aerial images has attracted increasing attention, especially open-vocabulary aerial detection which is not restricted to predefined categories. Due to the scarcity of drone's viewpoint images and their significant differences from natural images, it is difficult to achieve satisfying results by directly applying vanilla open-vocabulary detection methods designed for natural scenarios. Some studies propose to transfer knowledge from pre-trained models by using lightweight networks or generating pseudo labels, but they tend to rely on models trained on natural images, neglecting the potential of foundation models specifically tailored for remote sensing and aerial imagery. To address this limitation, we propose DisDop, a unified framework that systematically distills multi-level domain priors from remote sensing foundation models (e.g., RemoteCLIP and DINOv3) into a lightweight detector. Specifically, we first distill visual priors through a teacher fusion strategy that combines RemoteCLIP's cross-modal alignment capability with DINOv3's fine-grained local feature extraction ability, transferring their complementary strengths to the detector's backbone. Second, we distill textual priors embedded in RemoteCLIP's text encoder by explicitly modeling inter-category semantic relationships, while incorporating global contextual priors to enhance local feature representation for small objects. Through this multi-level prior distillation framework, our DisDop achieves new state-of-the-art performance on open-vocabulary aerial detection benchmarks. Extensive ablation analysis also demonstrates the rationality and effectiveness of our proposed modules.

AME-TS: Anchored Mixture-of-Experts for Time Series Forecasting

✇cs.AI, q-bio.NC updates on arXiv.org
By: Rui Wang · Renhao Xue · Ray Razi · Huan Song · Hannah R. Marlowe
26 May 2026 at 12:00
arXiv:2605.25166v1 Announce Type: cross Abstract: Time series forecasting models are increasingly scaled through large Transformer backbones, yet most existing approaches process all series through a shared dense computation path despite substantial heterogeneity in temporal structure. Mixture-of-Experts (MoE) offers a natural alternative by enabling conditional computation, but standard MoE routing leaves expert specialization weakly identified and often unstable during downstream adaptation. We propose AME-TS, a structure-guided sparse time series foundation model that aligns expert routing with interpretable temporal structure. AME-TS first uses a lightweight regime predictor to estimate series-level descriptors, including forecastability, seasonality, trend, and sparsity, and maps them to a soft structural prior over experts. This series-level prior guides token-level routing during training, encouraging structure-aligned specialization. On the GIFT-Eval benchmark, AME-TS delivers a strong accuracy-efficiency tradeoff across model scales: it substantially outperforms existing time series foundation models at small model scales and remains competitive with the strongest models at larger scales, while activating substantially fewer parameters through sparse routing. We further show that AME-TS learns more interpretable routing geometry and substantially more stable expert specialization than standard MoE during fine-tuning on the M5 dataset. These results suggest that structure-aware routing is an effective and reliable way to realize the benefits of sparse expert models for time series forecasting.

BC Protocol: Structured Dual-Expert Dialogue for Eliciting High-Quality Chain-of-Thought Post-Training Data

✇cs.AI, q-bio.NC updates on arXiv.org
By: Bo Zou · Chao Xu
26 May 2026 at 12:00
arXiv:2605.25549v1 Announce Type: cross Abstract: High-quality expert chain-of-thought (CoT) data is one of the core bottlenecks in large language model (LLM) post-training. Existing data production methods each have structural limitations: crowdsourced annotation lacks deep reasoning paths; expert solo writing is constrained by the "expert blind spot" -- experts structurally skip reasoning steps they consider obvious; RLHF only produces preference signals rather than reasoning chains. This paper proposes the BC Protocol -- a structured dual-expert elicitation method for LLM post-training data production. The method carefully pairs a domain expert (crystallized intelligence) with a knowledge engineer (fluid intelligence), systematically externalizing the expert's implicit judgments as natural language reasoning chains. We introduce the Participant Aptitude Model, which defines six participant characteristic dimensions that affect elicitation quality. "Calibrated Ignorance" is an original concept proposed in this paper. We further propose "Selection-over-Prescription" as a methodological principle: for implicit knowledge elicitation tasks, investing quality-control resources in personnel selection yields a higher return than investing the same resources in process design. In a controlled experiment in the narrative fiction domain, we directly compared CoT produced by BC Protocol dual dialogue (Group A, (n=20)) against CoT written independently by the same domain expert (Group B, (n=20)). Three cross-vendor judge models -- GPT-4o, Claude Opus 4.5, and Gemini 2.5 Pro -- conducted blind evaluation across five dimensions (600 ratings total). Results show that the BC Protocol achieves an overwhelming advantage in "naturalness of reasoning process" (Group A mean 4.80 vs. Group B mean 1.30, (p=2.4\times10^{-8}), Cliff's (\delta=1.0)).

When Search Becomes Memory: Turning Robot Design Trials into Transferable Skills

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yunfei Wang · Xiaohao Xu · Yang Li · Xiaonan Huang
26 May 2026 at 12:00
arXiv:2605.25832v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used as proposal generators for evolutionary robot design, yet most loops remain memoryless: simulator results shape the next population but are not preserved as reusable design knowledge. We present Auto-Robotist, a self-evolving LLM agent that distills morphology-search traces into an explicit natural-language skill library. Each skill stores a structural archetype, evidence-grounded positive and negative rules, and the evaluated designs that support them, making design memory inspectable rather than implicit in a population. During search, the agent retrieves skills to condition LLM edits of elite bodies while retaining a Genetic Algorithm (GA) mutation path for exploration; after evaluation, it updates the library through Add, Diagnose, and Merge. Across seven EvoGym tasks spanning locomotion, traversal, and object interaction, Auto-Robotist improves cold-start 5x5 search and transfers learned skills to 10x10 design spaces, where reference-conditioned transfer outperforms GA on every task. These results suggest that LLM agents can convert expensive physical evaluations into reusable, auditable design principles. Our code will be released upon acceptance.

QUIET: A Multi-Blank Cascaded Story Cloze Benchmark for LLM Creative Generation Capability

✇cs.AI, q-bio.NC updates on arXiv.org
By: Bo Zou · Chao Xu
26 May 2026 at 12:00
arXiv:2605.25955v1 Announce Type: cross Abstract: Large language models (LLMs) face a dual challenge in creative capability evaluation: existing benchmarks (e.g., Story Cloze Test, HellaSwag) measure models' discriminative ability over narrative continuation using multiple-choice recognition paradigms, rather than directly measuring creative generation capability; rubric-based scoring and LLM-as-Judge methods rely on subjective dimension assessment or natural language model outputs, and cannot provide objective, automated scoring mechanisms. This paper proposes QUIET (Quality Understanding via Interlocked Evaluation Testing), a diagnostic benchmark for LLM creative capability based on multi-blank cascaded story cloze. QUIET sets N blanks (10-20) in a story with complete structure, with each blank accompanied by an explicit content constraint, and cascade dependency relationships between blanks -- the content filled into earlier blanks constrains the feasible solution space for later blanks. The evaluated model (or human participants) fills all blanks in open-ended generation mode; the results are scored by an information-theoretic automated scoring protocol without human grading. The scoring protocol directly operationalizes the "calibrated surprise" theoretical framework (Zou & Xu, 2026a). For each blank k, a composite score is computed: score = satisfy * (1 + lambda * surprise), where lambda = 1.0. Here, "satisfy" measures how well the blank filling satisfies the content constraint (objective logical reasoning judgment, not subjective aesthetic scoring), and "surprise" measures the degree of surprise given that the constraint is satisfied. Creative answers that do not satisfy the constraint score zero; answers that satisfy the constraint but are mediocre score low; answers that satisfy the constraint and are surprising score high.

Creative Quality Alignment: Expert Tacit Knowledge Transfer via Chain-of-Thought Fine-Tuning

✇cs.AI, q-bio.NC updates on arXiv.org
By: Bo Zou · Chao Xu
26 May 2026 at 12:00
arXiv:2605.25977v1 Announce Type: cross Abstract: This paper provides an empirical implementation of the creative quality metric proposed in Calibrated Surprise (Zou & Xu, 2026a). The question this paper addresses is: does this mathematical claim hold at the engineering level? To make the answer as general as possible, we deliberately choose the strictest engineering conditions: low data cost and a small base model. Training data comes from approximately 100 expert chain-of-thought (CoT) annotations produced by the BC Protocol (Zou & Xu, 2026b). We also identify a data bias: most publicly available alignment datasets are skewed toward craft-related knowledge, while audience modeling and reality-logic coverage are systematically weak. We use the term Creative Quality Alignment (CQA) to describe this class of engineering methods. We also offer a supporting theoretical observation: in an LLM with a single conditional distribution architecture, calibrating the appreciation side automatically transfers to the generation side via architectural duality. This is the structural reason why ~100 CoT examples are sufficient -- not a purely empirical observation like LIMA (Zhou et al., 2023).

AnyMo: Geometry-Aware Setup-Agnostic Modeling of Human Motion in the Wild

✇cs.AI, q-bio.NC updates on arXiv.org
By: Baiyu Chen · Zechen Li · Wilson Wongso · Lihuan Li · Xiachong Lin · Hao Xue · Benjamin Tag · Flora Salim
26 May 2026 at 12:00
arXiv:2605.22715v2 Announce Type: replace-cross Abstract: As wearable and mobile devices become increasingly embedded in daily life, they offer a practical way to continuously sense human motion in the wild. But inertial signals are highly dependent on the sensing setup, including body location, mounting position, sensor orientation, device hardware, and sampling protocol. This setup dependence makes it difficult to learn motion representations that transfer across devices and datasets, and limits the broader use of wearable IMUs beyond closed-set recognition. We introduce AnyMo, a geometry-aware framework for setup-agnostic human motion modeling. AnyMo uses physics-grounded IMU simulation over dense body-surface placements to generate diverse and plausible synthetic signals, pre-trains a graph encoder from paired synthetic placement views and masked partial observations, tokenizes multi-position IMU into full-body motion tokens, and aligns these tokens with an LLM for motion-language understanding. We evaluate AnyMo on three complementary tasks: zero-shot activity recognition across 14 unseen downstream datasets, cross-modal retrieval, and wearable IMU motion captioning, where it improves average Accuracy/F1/R@2 by 11.7\%/11.6\%/22.6\% on HAR, increases zero-shot IMU-to-text and text-to-IMU retrieval MRR by 15.9\% and 28.6\%, respectively, and improves zero-shot captioning BERT-F1 by 18.8\%. These results support AnyMo as a generalist model for wearable motion understanding in the wild. Project page: https://baiyuchen.com/project/AnyMo.
Received — 13 April 2026 ⏭

Clinical application of base editing for treating β-thalassaemia

✇Nature - Issue - nature.com science feeds
By: Yongrong Lai · Rongrong Liu · Lijie Wang · Xu-Kai Ma · Yaliang Li · Gaohui Yang · Lingling Shi · Yi-Lin Guo · Zhenbin Wei · Xuemei Zhou · Wenchao Xu · Yaofeng Hou · Annarita Miccio · Bei Yang · Xiaodun Mou · Li Yang · Jia Chen
8 April 2026 at 08:00

Nature, Published online: 08 April 2026; doi:10.1038/s41586-026-10342-9

A clinical phase 1 trial of a single infusion of CS-101, CD34+ cells modified using a transformer base editor to reactivate fetal haemoglobin production, led to early and enduring transfusion independence in patients with β-thalassaemia.
Received — 4 April 2026 ⏭

SelfGrader: Stable Jailbreak Detection for Large Language Models using Token-Level Logits

✇cs.AI, q-bio.NC updates on arXiv.org
By: Zikai Zhang · Rui Hu · Olivera Kotevska · Jiahao Xu
3 April 2026 at 12:00
arXiv:2604.01473v1 Announce Type: cross Abstract: Large Language Models (LLMs) are powerful tools for answering user queries, yet they remain highly vulnerable to jailbreak attacks. Existing guardrail methods typically rely on internal features or textual responses to detect malicious queries, which either introduce substantial latency or suffer from the randomness in text generation. To overcome these limitations, we propose SelfGrader, a lightweight guardrail method that formulates jailbreak detection as a numerical grading problem using token-level logits. Specifically, SelfGrader evaluates the safety of a user query within a compact set of numerical tokens (NTs) (e.g., 0-9) and interprets their logit distribution as an internal safety signal. To align these signals with human intuition of maliciousness, SelfGrader introduces a dual-perspective scoring rule that considers both the maliciousness and benignness of the query, yielding a stable and interpretable score that reflects harmfulness and reduces the false positive rate simultaneously. Extensive experiments across diverse jailbreak benchmarks, multiple LLMs, and state-of-the-art guardrail baselines demonstrate that SelfGrader achieves up to a 22.66% reduction in ASR on LLaMA-3-8B, while maintaining significantly lower memory overhead (up to 173x) and latency (up to 26x).
Received — 2 April 2026 ⏭

Surgery-centered integrated strategies for personalized hepatocellular carcinoma care

✇MRD
By: Yuhan Cui · Zuchao Du · Jiahao Xu · Lihui Gu · Jiaqi Du · Mingda Wang · Tian Yang
31 March 2026 at 18:00

Cancer Biol Med. 2026 Mar 30:j.issn.2095-3941.2026.0045. doi: 10.20892/j.issn.2095-3941.2026.0045. Online ahead of print.

ABSTRACT

Hepatocellular carcinoma (HCC) remains a major global health burden characterized by late-stage diagnosis and high postoperative recurrence rates. This review presents a surgery-centered precision management framework integrating 3 synergistic components: early detection, precision surgery, and recurrence prevention. Early detection strategies incorporate multiparameter risk models including the gender, age, AFP-L3, AFP, and DCP (GALAD) as well as age, sex, AFP, and PIVKA-II (ASAP) scores, alongside circulating tumor DNA methylation-based liquid biopsy, thus enabling tumor identification at stages amenable to curative resection. Precision surgery optimizes patient selection through refined staging systems including the Chinese liver cancer staging (CNLC), and functional assessments including the albumin-bilirubin (ALBI) grade, whereas conversion therapy and minimally invasive approaches extend surgical eligibility to selected patients with intermediate-stage disease. To mitigate the risk of postoperative recurrence, distinguishing between early and late recurrence patterns and monitoring minimal residual disease are critical strategies. Perioperative systemic therapies, particularly immune checkpoint inhibitor-based combinations, show promise for eradicating micrometastatic disease. This integrated framework provides a cohesive, evidence-based approach to personalized HCC management aimed at maximizing curative potential and long-term survival.

PMID:41913379 | DOI:10.20892/j.issn.2095-3941.2026.0045

Received — 26 March 2026 ⏭

Parasites trigger epithelial cell crosstalk to drive gut–brain signalling

✇Nature - Issue - nature.com science feeds
By: Kouki K. Touhara · Jinhao Xu · Joel Castro · Hong-Erh Liang · Guochuan Li · Mariana Brizuela · Andrea M. Harrington · Sonia Garcia-Caraballo · Tracey O’Donnell · Daniel Neumann · Nathan D. Rossen · Fei Deng · Gudrun Schober · Yulong Li · Richard M. Locksley · Stuart M. Brierley · David Julius
25 March 2026 at 08:00

Nature, Published online: 25 March 2026; doi:10.1038/s41586-026-10281-5

Paracrine signalling between tuft cells and enterochromaffin cells is a key mode of immune–sensory and gut–brain communication, and accounts for the pattern of gastrointestinal symptoms that occurs during parasite infections.

Single-cell multiomics uncovers an endothelial mechanosensitive PIEZO1-IL-33 axis driving pulmonary fibrosis

✇Omics In Lung
By: Lanlan Zhang · Xuezhen Gui · Ruijie Hou · Liping Jia · Shu Xia · Xin Zhang · Yingyun Fu · Qian-Fang Meng · Qun Luo · Xing Shi · Bingxin Guo · Ruifang Liang · Ludan Yue · Xue Chen · Haizhao Xu · Pengbo Wang · Xia Tong · Lujie Liu · Lingwei Wang · Baicun Li · Zi Chen · Linfu Zhou · Linshu Zhang · Rongchang Chen · Changbin Sun · Wei Xu · Lang Rao · Haibo Zhou · Bi-Sen Ding · Shanze Chen
21 March 2026 at 18:00

Nat Commun. 2026 Mar 20;17(1):2655. doi: 10.1038/s41467-026-70193-w.

ABSTRACT

Pulmonary fibrosis represents a progressive interstitial lung disease marked by excessive extracellular matrix deposition and architectural distortion. Vascular endothelial cells critically contribute to fibrogenesis through paracrine secretion of pro-fibrotic mediators, yet their mechanobiological regulation remains elusive. Using integrated single-cell multi-omics profiling of human pulmonary fibrosis specimens and experimental fibrosis models induced by bleomycin or silica, we identify mechanosensitive Piezo1 upregulation in Endothelial cells as a hallmark of fibrotic progression. Endothelial-specific Piezo1 knockout significantly attenuates Bleomycin-induced fibrotic remodeling in male mice, establishing its pathogenic necessity. Mechanistically, PIEZO1 activation promotes pulmonary fibrosis development via CAPN2-mediated STAT3 phosphorylation, which may regulate the secretion of the pro-fibrotic molecule interleukin-33. These findings suggest that the endothelial PIEZO1-CAPN2-STAT3-IL33 axis is a potential therapeutic target for PF intervention.

PMID:41862476 | PMC:PMC13004862 | DOI:10.1038/s41467-026-70193-w

Received — 17 March 2026 ⏭

Towards unified brain-to-text decoding across speech production and perception

✇cs.AI, q-bio.NC updates on arXiv.org
By: Zhizhang Yuan · Yang Yang · Gaorui Zhang · Baowen Cheng · Zehan Wu · Yuhao Xu · Xiaoying Liu · Liang Chen · Ying Mao · Meng Li
16 March 2026 at 12:00
arXiv:2603.12628v1 Announce Type: new Abstract: Speech production and perception are the main ways humans communicate daily. Prior brain-to-text decoding studies have largely focused on a single modality and alphabetic languages. Here, we present a unified brain-to-sentence decoding framework for both speech production and perception in Mandarin Chinese. The framework exhibits strong generalization ability, enabling sentence-level decoding when trained only on single-character data and supporting characters and syllables unseen during training. In addition, it allows direct and controlled comparison of neural dynamics across modalities. Mandarin speech is decoded by first classifying syllable components in Hanyu Pinyin, namely initials and finals, from neural signals, followed by a post-trained large language model (LLM) that maps sequences of toneless Pinyin syllables to Chinese sentences. To enhance LLM decoding, we designed a three-stage post-training and two-stage inference framework based on a 7-billion-parameter LLM, achieving overall performance that exceeds larger commercial LLMs with hundreds of billions of parameters or more. In addition, several characteristics were observed in Mandarin speech production and perception: speech production involved neural responses across broader cortical regions than auditory perception; channels responsive to both modalities exhibited similar activity patterns, with speech perception showing a temporal delay relative to production; and decoding performance was broadly comparable across hemispheres. Our work not only establishes the feasibility of a unified decoding framework but also provides insights into the neural characteristics of Mandarin speech production and perception. These advances contribute to brain-to-text decoding in logosyllabic languages and pave the way toward neural language decoding systems supporting multiple modalities.

CRAFT-GUI: Curriculum-Reinforced Agent For GUI Tasks

✇cs.AI, q-bio.NC updates on arXiv.org
By: Songqin Nong · Xiaoxuan Tang · Jingxuan Xu · Sheng Zhou · Jianfeng Chen · Tao Jiang · Wenhao Xu
16 March 2026 at 12:00
arXiv:2508.11360v2 Announce Type: replace Abstract: As autonomous agents become adept at understanding and interacting with graphical user interface (GUI) environments, a new era of automated task execution is emerging. Recent studies have demonstrated that Reinforcement Learning (RL) can effectively enhance agents' performance in dynamic interactive GUI environments. However, these methods face two key limitations: (1) they overlook the significant variation in difficulty across different GUI tasks by treating the entire training data as a uniform set, which hampers the agent's ability to adapt its learning process; and (2) most approaches collapse task-specific nuances into a single, coarse reward, leaving the agent with a uniform signal that yields inefficient policy updates. To address these limitations, we propose CRAFT-GUI, a curriculum learning framework based on Group Relative Policy Optimization (GRPO) that explicitly accounts for the varying difficulty across trajectories. To enable more fine-grained policy optimization, we design a reward function that combines simple rule-based signals with model-judged evaluation, providing richer and more nuanced feedback during training. Experimental results demonstrate that our method achieves significant improvements over previous state-of-the-art approaches, outperforming them by 5.6% on public benchmarks Android Control and 10.3% on our internal online benchmarks, respectively. These findings empirically validate the effectiveness of integrating reinforcement learning with curriculum learning in GUI interaction tasks.
Received — 11 March 2026 ⏭

Fibration Policy Optimization

✇cs.AI, q-bio.NC updates on arXiv.org
By: Chang Li · Tshihao Tsu · Yaren Zhang · Chao Xue · Xiaodong He
10 March 2026 at 12:00
arXiv:2603.08239v1 Announce Type: cross Abstract: Large language models are increasingly trained as heterogeneous systems spanning multiple domains, expert partitions, and agentic pipelines, yet prevalent proximal objectives operate at a single scale and lack a principled mechanism for coupling token-level, trajectory-level, and higher-level hierarchical stability control. To bridge this gap, we derive the Aggregational Policy Censoring Objective (APC-Obj), the first exact unconstrained reformulation of sample-based TV-TRPO, establishing that clipping-based surrogate design and trust-region optimization are dual formulations of the same problem. Building on this foundation, we develop Fiber Bundle Gating (FBG), an algebraic framework that organizes sampled RL data as a fiber bundle and decomposes ratio gating into a base-level gate on trajectory aggregates and a fiber-level gate on per-token residuals, with provable first-order agreement with the true RL objective near on-policy. From APC-Obj and FBG we derive Fibration Policy Optimization (or simply, FiberPO), a concrete objective whose Jacobian is block-diagonal over trajectories, reduces to identity at on-policy, and provides better update direction thus improving token efficiency. The compositional nature of the framework extends beyond the trajectory-token case: fibrations compose algebraically into a Fibration Gating Hierarchy (FGH) that scales the same gating mechanism to arbitrary hierarchical depth without new primitives, as demonstrated by FiberPO-Domain, a four-level instantiation with independent trust-region budgets at the domain, prompt group, trajectory, and token levels. Together, these results connect the trust-region theory, a compositional algebraic structure, and practical multi-scale stability control into a unified framework for LLM policy optimization.

Oscillatory shear stress-driven endothelial-to-mesenchymal transition: a critical mechanical signal transduction mechanism in atherosclerosis progression

✇Cell Death Discovery nature.com science feeds
By: Jie Li · Wenchao Xu · Jie Ju · Min Liu · Wenxu Wang · Min Cheng · Xiaoyun Zhang · Xiaodong Cui · Hao Chen
10 March 2026 at 08:00

Cell Death Discovery, Published online: 10 March 2026; doi:10.1038/s41420-026-03000-6

Oscillatory shear stress-driven endothelial-to-mesenchymal transition: a critical mechanical signal transduction mechanism in atherosclerosis progression
Received — 5 March 2026 ⏭

MIND: Unified Inquiry and Diagnosis RL with Criteria Grounded Clinical Supports for Psychiatric Consultation

✇cs.AI, q-bio.NC updates on arXiv.org
By: Guoyi Li · Shihao Xu · Jiatong Ma · Yunyun Han · Jianhua Chen · Yafeng Deng
5 March 2026 at 13:00
arXiv:2603.03677v1 Announce Type: cross Abstract: Large language models (LLMs) have advanced medical dialogue systems, yet psychiatric consultation poses substantially higher demands due to subjective ambiguity and comorbidity complexity: an agent must continuously extract psychopathological cues from incomplete and inconsistent patient reports in multi-turn interactions and perform rigorous differential diagnostic reasoning. However, existing methods face two fundamental challenges. First, without criteria-grounded clinical supports, they are prone to unsupported clinical assertions when symptoms are atypical or underspecified. Second, in multi-turn interactions, they struggle to mitigate inquiry drift (off-topic or low-yield questioning) and optimize questioning strategies. To address these challenges, we propose MIND, a unified inquiry--diagnosis reinforcement learning framework for psychiatric consultation. Specifically, we build a Criteria-Grounded Psychiatric Reasoning Bank (PRB) that summarizes dialogue context into clinical retrieval states, retrieves semantically similar reference consultations, and distills reusable criteria-grounded clinical supports to guide criteria-aligned inquiry and reasoning. Building on this foundation, MIND enforces explicit clinical reasoning with rubric-based process rewards to provide fine-grained supervision over intermediate decision steps, and incorporates a value-aware trajectory rectification mechanism to jointly improve information acquisition and diagnostic decision-making across turns. Extensive experiments demonstrate that MIND consistently outperforms strong baselines in diagnostic accuracy, empathetic interaction quality, interpretability, and generalization.
❌