cs.AI, q-bio.NC updates on arXiv.org
CODESKILL: Learning Self-Evolving Skills for Coding Agents Yanzhou Li · Yiran Zhang · Xiaoyu Zhang · Xiaoxia Liu · Yang Liu
arXiv:2605.25430v1 Announce Type: new Abstract: Coding agents produce rich trajectories while solving software-engineering tasks. To enable agent self-evolution, these trajectories can be distilled into reusable procedural skills that compactly encode experience to guide future behavior. However, existing skill construction and maintenance methods often rely on fixed prompts and heuristic update rules, leaving it unclear how knowledge should be selected, abstracted, and maintained to best serve
26 May 2026 at 12:00

CODESKILL: Learning Self-Evolving Skills for Coding Agents

By: Yanzhou Li · Yiran Zhang · Xiaoyu Zhang · Xiaoxia Liu · Yang Liu

26 May 2026 at 12:00

arXiv:2605.25430v1 Announce Type: new Abstract: Coding agents produce rich trajectories while solving software-engineering tasks. To enable agent self-evolution, these trajectories can be distilled into reusable procedural skills that compactly encode experience to guide future behavior. However, existing skill construction and maintenance methods often rely on fixed prompts and heuristic update rules, leaving it unclear how knowledge should be selected, abstracted, and maintained to best serve downstream agents. We propose CODESKILL, an LLM-based framework that reformulates skill extraction and skill-bank maintenance as a learnable management policy. CODESKILL extracts multi-granularity procedural skills from coding-agent trajectories, evolves skills with new experience, and maintains a compact skill bank for future task solving. We train CODESKILL with reinforcement learning, using a hybrid reward that combines dense rubric-based skill-quality feedback with sparse verifiable execution feedback from the frozen downstream agent. Experiments on EnvBench, SWE-Bench Verified, and Terminal-Bench 2 show that CODESKILL improves average pass rate by 9.69 over the no-skill baseline and by 4.01 over the strongest prompt-based or memory baseline, while maintaining the skill bank at a stable size during iterative construction.

cs.AI, q-bio.NC updates on arXiv.org
What Are We Actually Decoding? Source Attribution for Non-Invasive Brain-to-Language Retrieval Xinyu Zhang · Sichao Liu · Runhao Lu · Alexandra Woolgar · Lihui Wang
arXiv:2605.24524v1 Announce Type: cross Abstract: In non-invasive neural language decoding, results can be inflated by sources that are not stimulus-evoked neural evidence: decoder priors, embedding-based metrics, and non-neural structural nuisances such as signal duration. The methodological challenge is therefore attribution: a reported gain is more informative when it can be traced to a specific source. We recast stimulus-locked MEG-to-audio retrieval as an auditing framework that separates
26 May 2026 at 12:00

What Are We Actually Decoding? Source Attribution for Non-Invasive Brain-to-Language Retrieval

cs.AI, q-bio.NC updates on arXiv.org

By: Xinyu Zhang · Sichao Liu · Runhao Lu · Alexandra Woolgar · Lihui Wang

26 May 2026 at 12:00

arXiv:2605.24524v1 Announce Type: cross Abstract: In non-invasive neural language decoding, results can be inflated by sources that are not stimulus-evoked neural evidence: decoder priors, embedding-based metrics, and non-neural structural nuisances such as signal duration. The methodological challenge is therefore attribution: a reported gain is more informative when it can be traced to a specific source. We recast stimulus-locked MEG-to-audio retrieval as an auditing framework that separates apparent performance into three sources - structural shortcuts, window-level stimulus-locked evidence, and cross-window contextual aggregation - and provides a diagnostic for each. Signal-blind Gaussian noise reaches 66.3% Rank@1 (R@1) under variable-length decoding but collapses to near chance once fixed-duration windows and stimulus-identity splits are enforced, isolating structural leakage. Under these controls, fixed-window retrieval recovers measurable MEG-audio discriminability, while an oracle sentence-bucket diagnostic shows that 95.7% of Top-1 errors select the wrong sentence, localising the residual bottleneck to sentence-level competition. We audit this contextual source with Group Context Bias (GCB), an inference-time additive logit bias that pools sentence-consistent evidence across windows while leaving the base retrieval scores and candidate pool fixed. Used as a score-space intervention, GCB makes the contextual source measurable: R@1 shifts from 44% to 52% on Gwilliams and from 22% to 29% on MOUS under the same fixed setting. GCB is auditable under this design: its effect collapses under random-grouping perturbations and vanishes when local evidence is attenuated in MEG or is near chance in EEG, supporting its use as a controlled source-attribution intervention. These results suggest that brain-to-language performance should be source-attributed, not merely reported.

cs.AI, q-bio.NC updates on arXiv.org
Rethinking Federated Unlearning via the Lens of Memorization Jiaheng Wei · Yanjun Zhang · He Zhang · Leo Yu Zhang · Chao Chen · Kok-Leong Ong · Jun Zhang · Yang Xiang
arXiv:2605.24545v1 Announce Type: cross Abstract: Federated learning (FL) increasingly needs machine unlearning to comply with privacy regulations. However, existing federated unlearning approaches may overlook the overlapping information between the unlearning and remaining data, leading to ineffective unlearning and unfairness between clients. In this work, we revisit federated unlearning through the lens of memorization. We argue that unlearning should mainly remove the unique memorized info
26 May 2026 at 12:00

Rethinking Federated Unlearning via the Lens of Memorization

cs.AI, q-bio.NC updates on arXiv.org

By: Jiaheng Wei · Yanjun Zhang · He Zhang · Leo Yu Zhang · Chao Chen · Kok-Leong Ong · Jun Zhang · Yang Xiang

26 May 2026 at 12:00

arXiv:2605.24545v1 Announce Type: cross Abstract: Federated learning (FL) increasingly needs machine unlearning to comply with privacy regulations. However, existing federated unlearning approaches may overlook the overlapping information between the unlearning and remaining data, leading to ineffective unlearning and unfairness between clients. In this work, we revisit federated unlearning through the lens of memorization. We argue that unlearning should mainly remove the unique memorized information attributable to the data to be forgotten, while preserving overlapping patterns that are also supported by the remaining data. Specifically, we propose Grouped Memorization Evaluation, an example-level metric that separates memorized knowledge from overlapping knowledge. Building on this metric, we introduce Federated Memorization Pruning (FedMemPrune), a pruning-based unlearning approach that resets redundant parameters responsible for memorization. Extensive experiments show that FedMemPrune closely matches retraining-based unlearning baselines while more effectively eliminating memorization than existing federated unlearning algorithms, yielding strong unlearning performance without sacrificing the utility of retained knowledge.

cs.AI, q-bio.NC updates on arXiv.org
HoloFair: Unified T2I Fairness Evaluation and Fair-GRPO Debiasing Ruyi Chen · Lu Zhou · Xiaogang Xu · Chiyu Zhang · Jiafei Wu · Liming Fang
arXiv:2605.24687v1 Announce Type: cross Abstract: Text-to-Image (T2I) models have made significant strides in visual realism and semantic consistency, yet they often perpetuate and amplify societal biases. Existing evaluation methods typically address only single-dimensional biases, lacking perspectives to uncover model biases at social-related deeper semantic levels. We introduce HoloFair, a comprehensive benchmark framework for multidimensional demographic bias analysis. Built upon our large-
26 May 2026 at 12:00

HoloFair: Unified T2I Fairness Evaluation and Fair-GRPO Debiasing

cs.AI, q-bio.NC updates on arXiv.org

By: Ruyi Chen · Lu Zhou · Xiaogang Xu · Chiyu Zhang · Jiafei Wu · Liming Fang

26 May 2026 at 12:00

arXiv:2605.24687v1 Announce Type: cross Abstract: Text-to-Image (T2I) models have made significant strides in visual realism and semantic consistency, yet they often perpetuate and amplify societal biases. Existing evaluation methods typically address only single-dimensional biases, lacking perspectives to uncover model biases at social-related deeper semantic levels. We introduce HoloFair, a comprehensive benchmark framework for multidimensional demographic bias analysis. Built upon our large-scale fairness-oriented dataset and the SpaFreq (Spatial-Frequency) attribute classifier, this framework proposes the Multi-attribute, Group-wise Bias Index (MGBI) metric, designed to assess both intrinsic diversity and conditional biases. Beyond evaluation, we further introduce Fair-GRPO, a reinforcement-learning-based debiasing method that alters the distribution of generative models through a designed multi-objective reward function. E.g., experiments on the SD3.5-Medium model demonstrate that Fair-GRPO significantly improves multidimensional fairness while maintaining high image quality. We also analyze potential reward hacking phenomena and provide corresponding mitigation strategies. Code and dataset are available at https://github.com/1059684669/HoloFair

cs.AI, q-bio.NC updates on arXiv.org
Selective Test-Time Compute Scaling for Click-Through Rate Prediction via Uncertainty-Triggered Feature Path Exploration Moyu Zhang · Yun Chen · Yujun Jin · Jinxin Hu · Yu Zhang · Xiaoyi Zeng
arXiv:2605.24989v1 Announce Type: cross Abstract: Scaling test-time compute has proven highly effective for language models, yet this opportunity remains largely unexplored for industrial Click-Through Rate (CTR) prediction. CTR models suffer from a fundamental asymmetry: feature combinations well-represented in training yield confident predictions, while sparsely observed ones produce unreliable outputs. Existing training-phase solutions such as adaptive gating learn a fixed selection function
26 May 2026 at 12:00

Selective Test-Time Compute Scaling for Click-Through Rate Prediction via Uncertainty-Triggered Feature Path Exploration

cs.AI, q-bio.NC updates on arXiv.org

By: Moyu Zhang · Yun Chen · Yujun Jin · Jinxin Hu · Yu Zhang · Xiaoyi Zeng

26 May 2026 at 12:00

arXiv:2605.24989v1 Announce Type: cross Abstract: Scaling test-time compute has proven highly effective for language models, yet this opportunity remains largely unexplored for industrial Click-Through Rate (CTR) prediction. CTR models suffer from a fundamental asymmetry: feature combinations well-represented in training yield confident predictions, while sparsely observed ones produce unreliable outputs. Existing training-phase solutions such as adaptive gating learn a fixed selection function subject to the same sparsity, offering no per-instance recourse at deployment.We propose UTTSI (Uncertainty-Triggered Test-Time Selective Inference), a training-free model-agnostic framework that scales inference depth proportionally to per-instance uncertainty. A dual-signal estimator combining model logit confidence with a data-level frequency prior distinguishes epistemic uncertainty from aleatoric ambiguity. Every instance undergoes adaptive feature filtering to remove unreliable embeddings; uncertain instances additionally receive stochastic feature-path explorations whose predictions are aggregated via consistency-weighted ensembling. Confident instances bypass exploration entirely, keeping average overhead at approximately $2.8\times$ base model cost with worst-case latency unchanged.Experiments on four datasets with three backbone architectures demonstrate consistent, statistically significant gains over all training-phase baselines. A seven-day online A/B test further confirms a 5.3% relative CTR gain ($p

cs.AI, q-bio.NC updates on arXiv.org
Test-Time Self-Adaptive Conditioning for Stable Audio-Driven Talking-Head Generation Zhicheng Zhang · Lei Wang · Yu Zhang · Yongsheng Gao
arXiv:2605.25488v1 Announce Type: cross Abstract: Audio-driven talking-head generation has achieved remarkable progress with recent models such as AniTalker, FLOAT, and Sonic. Despite their success, most existing approaches rely on a single static reference image to condition the entire video generation process at inference stage. This static conditioning paradigm often creates a mismatch between fixed identity features and dynamically evolving facial motion, leading to identity drift, temporal
26 May 2026 at 12:00

Test-Time Self-Adaptive Conditioning for Stable Audio-Driven Talking-Head Generation

cs.AI, q-bio.NC updates on arXiv.org

By: Zhicheng Zhang · Lei Wang · Yu Zhang · Yongsheng Gao

26 May 2026 at 12:00

arXiv:2605.25488v1 Announce Type: cross Abstract: Audio-driven talking-head generation has achieved remarkable progress with recent models such as AniTalker, FLOAT, and Sonic. Despite their success, most existing approaches rely on a single static reference image to condition the entire video generation process at inference stage. This static conditioning paradigm often creates a mismatch between fixed identity features and dynamically evolving facial motion, leading to identity drift, temporal inconsistency, and degraded perceptual quality. We introduce Test-Time Self-Adaptive Conditioning (TT-SAC), a parameter-free inference framework that enables pretrained talking-head generators to adapt their conditioning representations during inference without retraining, gradient updates, or additional supervision. Instead of treating the reference portrait as immutable, TT-SAC composes the generator with its encoder in a feedback loop: the generator's own outputs are re-encoded to construct a refined conditioning representation that better aligns with the temporal dynamics of the synthesized sequence. A single adaptation step approximates a self-consistent equilibrium of the generative process, stabilizing identity and motion across time. We further provide theoretical analysis showing that test-time conditioning adaptation reduces feature variance and improves generative stability under mild Lipschitz assumptions, while exhibiting a principled bias-variance tradeoff that governs the optimal strength of adaptation. Extensive experiments on state-of-the-art talking-head generators and benchmark datasets demonstrate consistent improvements in lip-sync accuracy, temporal coherence, identity preservation, and perceptual fidelity. TT-SAC offers a model-agnostic and training-free strategy for enhancing generative video models, establishing test-time conditioning adaptation as an effective mechanism for stabilizing audio-driven portrait animation.

cs.AI, q-bio.NC updates on arXiv.org
IPR-1: Interactive Physical Reasoner Mingyu Zhang · Lifeng Zhuo · Tianxi Tan · Guocan Xie · Xian Nie · Yan Li · Renjie Zhao · Zizhu He · Ziyu Wang · Jiting Cai · Yong-Lu Li
arXiv:2511.15407v4 Announce Type: replace Abstract: Humans learn by observing, interacting with environments, and internalizing physics and causality. Here, we aim to ask whether an agent can similarly acquire human-like reasoning from interaction and keep improving with more experience. To study this, we introduce a Game-to-Unseen (G2U) benchmark of 1,000+ heterogeneous games that exhibit significant visual domain gaps. Existing approaches, including VLMs and world models, struggle to capture
26 May 2026 at 12:00

IPR-1: Interactive Physical Reasoner

cs.AI, q-bio.NC updates on arXiv.org

By: Mingyu Zhang · Lifeng Zhuo · Tianxi Tan · Guocan Xie · Xian Nie · Yan Li · Renjie Zhao · Zizhu He · Ziyu Wang · Jiting Cai · Yong-Lu Li

26 May 2026 at 12:00

arXiv:2511.15407v4 Announce Type: replace Abstract: Humans learn by observing, interacting with environments, and internalizing physics and causality. Here, we aim to ask whether an agent can similarly acquire human-like reasoning from interaction and keep improving with more experience. To study this, we introduce a Game-to-Unseen (G2U) benchmark of 1,000+ heterogeneous games that exhibit significant visual domain gaps. Existing approaches, including VLMs and world models, struggle to capture underlying physics and causality since they are not focused on core mechanisms and overfit to visual details. VLM/VLA agents reason but lack look-ahead in interactive settings, while world models imagine but imitate visual patterns rather than analyze physics and causality. We therefore propose IPR (Interactive Physical Reasoner), using world-model rollouts to score and reinforce a VLM's policy, and introduce PhysCode, a physics-centric action code aligning semantic intent with dynamics to provide a shared action space for prediction and reasoning. Pretrained on 1,000+ games, our IPR performs robustly on levels from primitive intuition to goal-driven reasoning, and even surpasses GPT-5 overall. We find that performance improves with more training games and interaction steps, and that the model also zero-shot transfers to unseen games. These results support physics-centric interaction as a path to steadily improving physical reasoning. Further demos and project details can be found at https://mybearyzhang.github.io/ipr-1.

cs.AI, q-bio.NC updates on arXiv.org
All Leaks Count, Some Count More: Interpretable Temporal Contamination Detection and Mitigation in LLM Backtesting Zeyu Zhang · Ryan Chen · Bradly C. Stadie
arXiv:2602.17234v2 Announce Type: replace Abstract: Backtesting LLMs on resolved events assumes models reason only from pre-cutoff knowledge, yet pretrained models inevitably leak post-cutoff knowledge. We introduce a claim-level evaluation framework that decomposes prediction rationales into atomic claims and applies Shapley values to quantify each claim's decision impact, yielding \textbf{Shapley-DCLR} (\textbf{Shapley}-weighted \textbf{D}ecision-\textbf{C}ritical \textbf{L}eakage \textbf{R}a
26 May 2026 at 12:00

All Leaks Count, Some Count More: Interpretable Temporal Contamination Detection and Mitigation in LLM Backtesting

cs.AI, q-bio.NC updates on arXiv.org

By: Zeyu Zhang · Ryan Chen · Bradly C. Stadie

26 May 2026 at 12:00

arXiv:2602.17234v2 Announce Type: replace Abstract: Backtesting LLMs on resolved events assumes models reason only from pre-cutoff knowledge, yet pretrained models inevitably leak post-cutoff knowledge. We introduce a claim-level evaluation framework that decomposes prediction rationales into atomic claims and applies Shapley values to quantify each claim's decision impact, yielding \textbf{Shapley-DCLR} (\textbf{Shapley}-weighted \textbf{D}ecision-\textbf{C}ritical \textbf{L}eakage \textbf{R}ate) -- an interpretable metric measuring what fraction of decision-driving reasoning is contaminated. We further propose \textbf{TimeSPEC} (\textbf{Time}-\textbf{S}upervised \textbf{P}rediction with \textbf{E}xtracted \textbf{C}laims), an inference-time architecture that interleaves temporally-filtered retrieval with claim-level supervision, producing predictions grounded entirely in pre-cutoff evidence. Across three LLMs, the ablation experiments confirm retrieval and supervision are jointly necessary; and a three-task probe further illstrates that the performance cost of temporal enforcement scales with each task's reliance on post-cutoff information.

cs.AI, q-bio.NC updates on arXiv.org
AutoResearchClaw: Self-Reinforcing Autonomous Research with Human-AI Collaboration Jiaqi Liu · Shi Qiu · Mairui Li · Bingzhou Li · Haonian Ji · Siwei Han · Xinyu Ye · Peng Xia · Zihan Dong · Meng Chen · Congyu Zhang · Letian Zhang · Guiming Chen · Haoqin Tu · Xinyu Yang · Lu Feng · Xujiang Zhao · Haifeng Chen · Jiawei Zhou · Xiao Wang · Weitong Zhang · Hongtu Zhu · Yun Li · Jieru Mei · Hongliang Fei · Jiaheng Zhang · Linjie Li · Linjun Zhang · Yuyin Zhou · Sheng Wang · Caiming Xiong · James Zou · Zeyu Zheng · Cihang Xie · Mingyu Ding · Huaxiu Yao
arXiv:2605.20025v2 Announce Type: replace Abstract: Automating scientific discovery requires more than generating papers from ideas. Real research is iterative: hypotheses are challenged from multiple perspectives, experiments fail and inform the next attempt, and lessons accumulate across cycles. Existing autonomous research systems often model this process as a linear pipeline: they rely on single-agent reasoning, stop when execution fails, and do not carry experience across runs. We present
26 May 2026 at 12:00

AutoResearchClaw: Self-Reinforcing Autonomous Research with Human-AI Collaboration

cs.AI, q-bio.NC updates on arXiv.org

By: Jiaqi Liu · Shi Qiu · Mairui Li · Bingzhou Li · Haonian Ji · Siwei Han · Xinyu Ye · Peng Xia · Zihan Dong · Meng Chen · Congyu Zhang · Letian Zhang · Guiming Chen · Haoqin Tu · Xinyu Yang · Lu Feng · Xujiang Zhao · Haifeng Chen · Jiawei Zhou · Xiao Wang · Weitong Zhang · Hongtu Zhu · Yun Li · Jieru Mei · Hongliang Fei · Jiaheng Zhang · Linjie Li · Linjun Zhang · Yuyin Zhou · Sheng Wang · Caiming Xiong · James Zou · Zeyu Zheng · Cihang Xie · Mingyu Ding · Huaxiu Yao

26 May 2026 at 12:00

arXiv:2605.20025v2 Announce Type: replace Abstract: Automating scientific discovery requires more than generating papers from ideas. Real research is iterative: hypotheses are challenged from multiple perspectives, experiments fail and inform the next attempt, and lessons accumulate across cycles. Existing autonomous research systems often model this process as a linear pipeline: they rely on single-agent reasoning, stop when execution fails, and do not carry experience across runs. We present AutoResearchClaw, a multi-agent autonomous research pipeline built on five mechanisms: structured multi-agent debate for hypothesis generation and result analysis, a self-healing executor with a \textsc{Pivot}/\textsc{Refine} decision loop that transforms failures into information, verifiable result reporting that prevents fabricated numbers and hallucinated citations, human-in-the-loop collaboration with seven intervention modes spanning full autonomy to step-by-step oversight, and cross-run evolution that converts past mistakes into future safeguards. On ARC-Bench, a 25-topic experiment-stage benchmark, AutoResearchClaw outperforms AI Scientist v2 by 54.7%. A human-in-the-loop ablation across seven intervention modes reveals that precise, targeted collaboration at high-leverage decision points consistently outperforms both full autonomy and exhaustive step-by-step oversight. We position AutoResearchClaw as a research amplifier that augments rather than replaces human scientific judgment. Code is available at https://github.com/aiming-lab/AutoResearchClaw.

Nature - Issue - nature.com science feeds
Advancing solar and wind penetration in China through energy complementarity Yuan Hu · Hou Jiang · Chuan Zhang · Jianlong Yuan · Mengting Zhang · Ling Yao · Qiang Chen · Jichao Wu · Hualong Zhang · Subin Ma · Xiang Li · Weiyu Zhang · Quanhua Dong · Congcong Wen · Gege Yin · Fan Zhang · Chaohui Yu · Zhijun Jin · Yu Liu
Nature, Published online: 20 May 2026; doi:10.1038/s41586-026-10570-zUsing high-resolution satellite imagery combined with a deep-learning-based framework to build a national energy inventory enables a data-driven assessment of solar–wind complementarity strategies to reduce power variability and enhance renewable energy penetration across China.
20 May 2026 at 08:00

Advancing solar and wind penetration in China through energy complementarity

Nature - Issue - nature.com science feeds

By: Yuan Hu · Hou Jiang · Chuan Zhang · Jianlong Yuan · Mengting Zhang · Ling Yao · Qiang Chen · Jichao Wu · Hualong Zhang · Subin Ma · Xiang Li · Weiyu Zhang · Quanhua Dong · Congcong Wen · Gege Yin · Fan Zhang · Chaohui Yu · Zhijun Jin · Yu Liu

20 May 2026 at 08:00

Nature, Published online: 20 May 2026; doi:10.1038/s41586-026-10570-z

Using high-resolution satellite imagery combined with a deep-learning-based framework to build a national energy inventory enables a data-driven assessment of solar–wind complementarity strategies to reduce power variability and enhance renewable energy penetration across China.

Nature - Issue - nature.com science feeds
EBV strain interacts with host HLA to drive nasopharyngeal carcinoma risk Yanhong Chen · Jingtong Liang · Wanlin Zhang · Xinyu Zhang · Xinyi Zhang · Ching-Yuan Wang · Fei Yao · Shanshan Zhang · Xiang Zhou · Weimin Ye · Ruimei Feng · Yonglin Cai · Zhe Zhang · Mingfang Ji · Qian Cui · Xihong Lin · Jiesen Li · Jialei Xu · Qiuting Zhang · Qinyao Huang · Yingying Cheng · Yanran Luo · Xiaoping Ye · Qisheng Feng · Minzhong Tang · Mu-Sheng Zeng · Yi-Xin Zeng · Zhonghua Liu · Weiwei Zhai · Jianjun Liu · Miao Xu
Nature, Published online: 15 April 2026; doi:10.1038/s41586-026-10416-8A genome-to-genome association study identifies host and viral risk factors that interact to drive nasopharyngeal carcinoma endemicity in southern China.
15 April 2026 at 08:00

EBV strain interacts with host HLA to drive nasopharyngeal carcinoma risk

Nature - Issue - nature.com science feeds

By: Yanhong Chen · Jingtong Liang · Wanlin Zhang · Xinyu Zhang · Xinyi Zhang · Ching-Yuan Wang · Fei Yao · Shanshan Zhang · Xiang Zhou · Weimin Ye · Ruimei Feng · Yonglin Cai · Zhe Zhang · Mingfang Ji · Qian Cui · Xihong Lin · Jiesen Li · Jialei Xu · Qiuting Zhang · Qinyao Huang · Yingying Cheng · Yanran Luo · Xiaoping Ye · Qisheng Feng · Minzhong Tang · Mu-Sheng Zeng · Yi-Xin Zeng · Zhonghua Liu · Weiwei Zhai · Jianjun Liu · Miao Xu

15 April 2026 at 08:00

Nature, Published online: 15 April 2026; doi:10.1038/s41586-026-10416-8

A genome-to-genome association study identifies host and viral risk factors that interact to drive nasopharyngeal carcinoma endemicity in southern China.

Omics In Lung
GPNMB Drives Brain Metastasis by Sculpting a Pathological Endothelial-Immune Interactome Xuefei Liu · Jun Tan · Chun Wu · Guanyin Huang · Yixin Cheng · Jianyang Hu · Binyu Zhang · Mao Zhao · Boxi Zhao · Jingru Lian · Shuqian Zheng · Lin Zeng · Meng Xu · Yang Xu · Shan Zeng · Hao Yu · Hui Yang · Zhixiang Zuo · Chuanyu Liu · Weineng Feng · Weinan Guo · Chunying Li · Sai-Lan Liu · Qing Liu · Feiqiu Wen · Xin Hong
Cancer Discov. 2026 Apr 15. doi: 10.1158/2159-8290.CD-25-1663. Online ahead of print.ABSTRACTBrain metastases (BM) remain a devastating disease with dismal prognosis. How circulating tumor cells (CTCs) penetrate the blood brain barrier (BBB) and reprogram the brain microenvironment remain unclear. Using spatially resolved multi-omic profiling of CTCs and brain metastases, integrated with experimental and clinical analyses, we identified Glycoprotein Non-Metastatic Melanoma Protein B (GPNMB) as a
13 April 2026 at 18:00

GPNMB Drives Brain Metastasis by Sculpting a Pathological Endothelial-Immune Interactome

Omics In Lung

By: Xuefei Liu · Jun Tan · Chun Wu · Guanyin Huang · Yixin Cheng · Jianyang Hu · Binyu Zhang · Mao Zhao · Boxi Zhao · Jingru Lian · Shuqian Zheng · Lin Zeng · Meng Xu · Yang Xu · Shan Zeng · Hao Yu · Hui Yang · Zhixiang Zuo · Chuanyu Liu · Weineng Feng · Weinan Guo · Chunying Li · Sai-Lan Liu · Qing Liu · Feiqiu Wen · Xin Hong

13 April 2026 at 18:00

Cancer Discov. 2026 Apr 15. doi: 10.1158/2159-8290.CD-25-1663. Online ahead of print.

ABSTRACT

Brain metastases (BM) remain a devastating disease with dismal prognosis. How circulating tumor cells (CTCs) penetrate the blood brain barrier (BBB) and reprogram the brain microenvironment remain unclear. Using spatially resolved multi-omic profiling of CTCs and brain metastases, integrated with experimental and clinical analyses, we identified Glycoprotein Non-Metastatic Melanoma Protein B (GPNMB) as a CTC-secreted driver of vascular disruption and brain colonization. CBX3 upregulation induced GPNMB expression, which bound endothelial EGFR, triggering CBL-mediated ubiquitination and degradation. Attenuated EGFR signaling suppressed FTO and disrupted endothelial junctions via YTHDF2-dependent TJP1 m6A methylation. Remarkably, GPNMB-induced BBB remodeling promoted immune infiltration via CXCL12-CXCR4 axis, and induced time course-dependent T cell exhaustion within the brain microenvironment. Clinically, elevated CBX3⁺GPNMB⁺ CTCs and plasma CXCL12 were significantly associated with BM progression in lung cancer and melanoma. Therapeutically, dual blockade of GPNMB and PD1 enhanced anti-BM efficacy in mice, unveiling GPNMB as a promising target for precision immunotherapy.

PMID:41973996 | DOI:10.1158/2159-8290.CD-25-1663

Omics In Lung
Applications and challenges of multi-omics approaches in lung cancer research and precision treatment Shouyu Zhang · Renjie Xu · Qian Zheng · Bojiang Chen · Weimin Li
Front Genet. 2026 Mar 23;16:1722368. doi: 10.3389/fgene.2025.1722368. eCollection 2025.ABSTRACTLung cancer is one of the most common cancers worldwide and one of the leading causes of cancer death, with a heavy disease burden and severe public health challenges. Multi-omics techniques, such as genomics, proteomics, metabolomics, and radiomics, play a crucial role in the early diagnosis and treatment of lung cancer, revealing the molecular characteristics and mechanisms of lung cancer, and have s
8 April 2026 at 18:00

Applications and challenges of multi-omics approaches in lung cancer research and precision treatment

Omics In Lung

By: Shouyu Zhang · Renjie Xu · Qian Zheng · Bojiang Chen · Weimin Li

8 April 2026 at 18:00

Front Genet. 2026 Mar 23;16:1722368. doi: 10.3389/fgene.2025.1722368. eCollection 2025.

ABSTRACT

Lung cancer is one of the most common cancers worldwide and one of the leading causes of cancer death, with a heavy disease burden and severe public health challenges. Multi-omics techniques, such as genomics, proteomics, metabolomics, and radiomics, play a crucial role in the early diagnosis and treatment of lung cancer, revealing the molecular characteristics and mechanisms of lung cancer, and have significant clinical application value. However, it also faces numerous challenges, such as data issues, "black box" problems, and ethical and legal concerns. How to leverage strengths while mitigating weaknesses, achieve clinical translation of technology, and serve patients more effectively deserves our deep reflection. This article reviews the specific applications and challenges of multi-omics methods in lung cancer research and personalized treatment.

PMID:41948518 | PMC:PMC13050793 | DOI:10.3389/fgene.2025.1722368

Nature - Issue - nature.com science feeds
Engineered immunosuppressive dendritic cells protect against cardiac remodelling Xiaoying Li · Jiamin Li · Guohua Li · Lisheng Zhu · Guo Cheng · Huanqiang Li · Hao Lin · Ningqing Jia · Xiaoqian Hong · Ye Liu · Zhiwei Zhong · Yize Chen · Biqing Wang · Jing Zhao · Zhenqi Hua · Lingjun Wang · Qiming Chen · Peijie Zheng · Shuyuan Sheng · Songting Gu · Cheng Ni · Shuchang Ye · Changle Ke · Feimu Zhang · Mo Li · Shaohui Shi · Junhua He · Yan Wu · Yinghui Xu · Minjian Kong · Qi Chen · Huajun Li · Yu Zhang · Jianzhong Sun · Guanhua Hu · Chengchen Zhao · Yiping Dong · Lili Yu · Yang Xu · Xinyang Hu
Nature, Published online: 08 April 2026; doi:10.1038/s41586-026-10346-5Lesion-targeted immune modulation is a feasible strategy to control cardiac fibrosis, and engineered dendritic cells are a promising therapeutic platform for treating cardiac remodelling and heart failure.
8 April 2026 at 08:00

Engineered immunosuppressive dendritic cells protect against cardiac remodelling

Nature - Issue - nature.com science feeds

By: Xiaoying Li · Jiamin Li · Guohua Li · Lisheng Zhu · Guo Cheng · Huanqiang Li · Hao Lin · Ningqing Jia · Xiaoqian Hong · Ye Liu · Zhiwei Zhong · Yize Chen · Biqing Wang · Jing Zhao · Zhenqi Hua · Lingjun Wang · Qiming Chen · Peijie Zheng · Shuyuan Sheng · Songting Gu · Cheng Ni · Shuchang Ye · Changle Ke · Feimu Zhang · Mo Li · Shaohui Shi · Junhua He · Yan Wu · Yinghui Xu · Minjian Kong · Qi Chen · Huajun Li · Yu Zhang · Jianzhong Sun · Guanhua Hu · Chengchen Zhao · Yiping Dong · Lili Yu · Yang Xu · Xinyang Hu

8 April 2026 at 08:00

Nature, Published online: 08 April 2026; doi:10.1038/s41586-026-10346-5

Lesion-targeted immune modulation is a feasible strategy to control cardiac fibrosis, and engineered dendritic cells are a promising therapeutic platform for treating cardiac remodelling and heart failure.

cs.AI, q-bio.NC updates on arXiv.org
Toward Executable Repository-Level Code Generation via Environment Alignment Ruwei Pan · Junlei Shen · Linhao Wu · Yueheng Zhu · Zixiong Yang · Yakun Zhang · Lu Zhang · Hongyu Zhang
arXiv:2604.03622v1 Announce Type: cross Abstract: Large language models (LLMs) have achieved strong performance on code generation, but existing methods still struggle with repository-level code generation under executable validation. Under this evaluation setting, success is determined not by the plausibility of isolated code fragments, but by whether a generated multi-file repository can be successfully installed, have its dependencies and internal references resolved, be launched, and be val
7 April 2026 at 12:00

Toward Executable Repository-Level Code Generation via Environment Alignment

cs.AI, q-bio.NC updates on arXiv.org

By: Ruwei Pan · Junlei Shen · Linhao Wu · Yueheng Zhu · Zixiong Yang · Yakun Zhang · Lu Zhang · Hongyu Zhang

7 April 2026 at 12:00

arXiv:2604.03622v1 Announce Type: cross Abstract: Large language models (LLMs) have achieved strong performance on code generation, but existing methods still struggle with repository-level code generation under executable validation. Under this evaluation setting, success is determined not by the plausibility of isolated code fragments, but by whether a generated multi-file repository can be successfully installed, have its dependencies and internal references resolved, be launched, and be validated in a real execution environment. To address this challenge, we propose EnvGraph, a framework for repository-level code generation that formulates repository executability as an environment alignment problem. EnvGraph jointly models two coupled conditions for successful repository execution, namely external dependency satisfaction and repository-internal reference resolution. It maintains a dual-layer environment representation, uses execution evidence to perform execution-evidence-based attribution, and guides repository generation through a unified targeted revision mechanism within an iterative alignment loop. We evaluate EnvGraph on repository-level code generation with three representative backbone LLMs and compare it against representative environment-aware and repository-level baselines. Experimental results show that EnvGraph consistently achieves the best performance on these repository-level benchmarks. In particular, it outperforms the strongest non-EnvGraph baseline by an absolute margin of 5.72--5.87 percentage points in Functional Correctness and 4.58--8.66 percentage points in Non-Functional Quality.

cs.AI, q-bio.NC updates on arXiv.org
Persistent Cross-Attempt State Optimization for Repository-Level Code Generation Ruwei Pan · Jiangshuai Wang · Qisheng Zhang · Yueheng Zhu · Linhao Wu · Zixiong Yang · Yakun Zhang · Lu Zhang · Hongyu Zhang
arXiv:2604.03632v1 Announce Type: cross Abstract: Large language models (LLMs) have achieved substantial progress in repository-level code generation. However, solving the same repository-level task often requires multiple attempts, while existing methods still optimize each attempt in isolation and do not preserve or reuse task-specific state across attempts. In this paper, we propose LiveCoder, a novel framework for repository-level code generation based on cross-attempt knowledge optimizatio
7 April 2026 at 12:00

Persistent Cross-Attempt State Optimization for Repository-Level Code Generation

cs.AI, q-bio.NC updates on arXiv.org

By: Ruwei Pan · Jiangshuai Wang · Qisheng Zhang · Yueheng Zhu · Linhao Wu · Zixiong Yang · Yakun Zhang · Lu Zhang · Hongyu Zhang

7 April 2026 at 12:00

arXiv:2604.03632v1 Announce Type: cross Abstract: Large language models (LLMs) have achieved substantial progress in repository-level code generation. However, solving the same repository-level task often requires multiple attempts, while existing methods still optimize each attempt in isolation and do not preserve or reuse task-specific state across attempts. In this paper, we propose LiveCoder, a novel framework for repository-level code generation based on cross-attempt knowledge optimization. LiveCoder maintains persistent task-specific state from prior attempts to guide subsequent generation. This state includes success knowledge, which captures reusable signals from previously strong repositories, failure knowledge, which records unsuccessful outcomes and their diagnostic signals, and a historical-best repository, which preserves the strongest result found so far and prevents regression. These components collectively transform repeated repository generation into a persistent, knowledge-driven optimization process. We evaluate LiveCoder using four frontier LLMs on two representative repository-level code generation benchmarks. Extensive experimental results demonstrate the effectiveness and efficiency of LiveCoder, improving the functional score by up to 22.94 percentage points, increasing repository reuse to 81.58%, and reducing cost by up to 53.63% on RAL-Bench while maintaining broadly stable non-functional quality.

cs.AI, q-bio.NC updates on arXiv.org
LightThinker++: From Reasoning Compression to Memory Management Yuqi Zhu · Jintian Zhang · Zhenjie Wan · Yujie Luo · Shuofei Qiao · Zhengke Gui · Da Zheng · Lei Liang · Huajun Chen · Ningyu Zhang
arXiv:2604.03679v1 Announce Type: cross Abstract: Large language models (LLMs) excel at complex reasoning, yet their efficiency is limited by the surging cognitive overhead of long thought traces. In this paper, we propose LightThinker, a method that enables LLMs to dynamically compress intermediate thoughts into compact semantic representations. However, static compression often struggles with complex reasoning where the irreversible loss of intermediate details can lead to logical bottlenecks
7 April 2026 at 12:00

LightThinker++: From Reasoning Compression to Memory Management

cs.AI, q-bio.NC updates on arXiv.org

By: Yuqi Zhu · Jintian Zhang · Zhenjie Wan · Yujie Luo · Shuofei Qiao · Zhengke Gui · Da Zheng · Lei Liang · Huajun Chen · Ningyu Zhang

7 April 2026 at 12:00

arXiv:2604.03679v1 Announce Type: cross Abstract: Large language models (LLMs) excel at complex reasoning, yet their efficiency is limited by the surging cognitive overhead of long thought traces. In this paper, we propose LightThinker, a method that enables LLMs to dynamically compress intermediate thoughts into compact semantic representations. However, static compression often struggles with complex reasoning where the irreversible loss of intermediate details can lead to logical bottlenecks. To address this, we evolve the framework into LightThinker++, introducing Explicit Adaptive Memory Management. This paradigm shifts to behavioral-level management by incorporating explicit memory primitives, supported by a specialized trajectory synthesis pipeline to train purposeful memory scheduling. Extensive experiments demonstrate the framework's versatility across three dimensions. (1) LightThinker reduces peak token usage by 70% and inference time by 26% with minimal accuracy loss. (2) In standard reasoning, LightThinker++ slashes peak token usage by 69.9% while yielding a +2.42% accuracy gain under the same context budget for maximum performance. (3) Most notably, in long-horizon agentic tasks, it maintains a stable footprint beyond 80 rounds (a 60%-70% reduction), achieving an average performance gain of 14.8% across different complex scenarios. Overall, our work provides a scalable direction for sustaining deep LLM reasoning over extended horizons with minimal overhead.

cs.AI, q-bio.NC updates on arXiv.org
Xpertbench: Expert Level Tasks with Rubrics-Based Evaluation Xue Liu · Xin Ma · Yuxin Ma · Yongchang Peng · Duo Wang · Zhoufutu Wen · Ge Zhang · Kaiyuan Zhang · Xinyu Chen · Tianci He · Jiani Hou · Liang Hu · Ziyun Huang · Yongzhe Hui · Jianpeng Jiao · Chennan Ju · Yingru Kong · Yiran Li · Mengyun Liu · Luyao Ma · Fei Ni · Yiqing Ni · Yueyan Qiu · Yanle Ren · Zilin Shi · Zaiyuan Wang · Wenjie Yue · Shiyu Zhang · Xinyi Zhang · Kaiwen Zhao · Zhenwei Zhu · Shanshan Wu · Qi Zhao · Wenhao Huang
arXiv:2604.02368v3 Announce Type: replace Abstract: As Large Language Models (LLMs) exhibit plateauing performance on conventional benchmarks, a pivotal challenge persists: evaluating their proficiency in complex, open-ended tasks characterizing genuine expert-level cognition. Existing frameworks suffer from narrow domain coverage, reliance on generalist tasks, or self-evaluation biases. To bridge this gap, we present XpertBench, a high-fidelity benchmark engineered to assess LLMs across authen
7 April 2026 at 12:00

Xpertbench: Expert Level Tasks with Rubrics-Based Evaluation

cs.AI, q-bio.NC updates on arXiv.org

By: Xue Liu · Xin Ma · Yuxin Ma · Yongchang Peng · Duo Wang · Zhoufutu Wen · Ge Zhang · Kaiyuan Zhang · Xinyu Chen · Tianci He · Jiani Hou · Liang Hu · Ziyun Huang · Yongzhe Hui · Jianpeng Jiao · Chennan Ju · Yingru Kong · Yiran Li · Mengyun Liu · Luyao Ma · Fei Ni · Yiqing Ni · Yueyan Qiu · Yanle Ren · Zilin Shi · Zaiyuan Wang · Wenjie Yue · Shiyu Zhang · Xinyi Zhang · Kaiwen Zhao · Zhenwei Zhu · Shanshan Wu · Qi Zhao · Wenhao Huang

7 April 2026 at 12:00

arXiv:2604.02368v3 Announce Type: replace Abstract: As Large Language Models (LLMs) exhibit plateauing performance on conventional benchmarks, a pivotal challenge persists: evaluating their proficiency in complex, open-ended tasks characterizing genuine expert-level cognition. Existing frameworks suffer from narrow domain coverage, reliance on generalist tasks, or self-evaluation biases. To bridge this gap, we present XpertBench, a high-fidelity benchmark engineered to assess LLMs across authentic professional domains. XpertBench consists of 1,346 meticulously curated tasks across 80 categories, spanning finance, healthcare, legal services, education, and dual-track research (STEM and Humanities). These tasks are derived from over 1,000 submissions by domain experts--including researchers from elite institutions and practitioners with extensive clinical or industrial experience--ensuring superior ecological validity. Each task uses detailed rubrics with mostly 15-40 weighted checkpoints to assess professional rigor. To facilitate scalable yet human-aligned assessment, we introduce ShotJudge, a novel evaluation paradigm that employs LLM judges calibrated with expert few-shot exemplars to mitigate self-rewarding biases. Our empirical evaluation of state-of-the-art LLMs reveals a pronounced performance ceiling: even leading models achieve a peak success rate of only ~66%, with a mean score around 55%. Models also exhibit domain-specific divergence, showing non-overlapping strengths in quantitative reasoning versus linguistic synthesis.. These findings underscore a significant "expert-gap" in current AI systems and establish XpertBench as a critical instrument for navigating the transition from general-purpose assistants to specialized professional collaborators.

cs.AI, q-bio.NC updates on arXiv.org
DWDP: Distributed Weight Data Parallelism for High-Performance LLM Inference on NVL72 Wanqian Li · Jintao Peng · Zongfei Jing · Tianyu Zhang · Ze Long · Xianjie Qiao · Xiaoming Chen · Dongxu Yang · Kefeng Duan · June Yang
arXiv:2604.01621v1 Announce Type: cross Abstract: Large language model (LLM) inference increasingly depends on multi-GPU execution, yet existing inference parallelization strategies require layer-wise inter-rank synchronization, making end-to-end performance sensitive to workload imbalance. We present DWDP (Distributed Weight Data Parallelism), an inference parallelization strategy that preserves data-parallel execution while offloading MoE weights across peer GPUs and fetching missing experts
3 April 2026 at 12:00

DWDP: Distributed Weight Data Parallelism for High-Performance LLM Inference on NVL72

cs.AI, q-bio.NC updates on arXiv.org

By: Wanqian Li · Jintao Peng · Zongfei Jing · Tianyu Zhang · Ze Long · Xianjie Qiao · Xiaoming Chen · Dongxu Yang · Kefeng Duan · June Yang

3 April 2026 at 12:00

arXiv:2604.01621v1 Announce Type: cross Abstract: Large language model (LLM) inference increasingly depends on multi-GPU execution, yet existing inference parallelization strategies require layer-wise inter-rank synchronization, making end-to-end performance sensitive to workload imbalance. We present DWDP (Distributed Weight Data Parallelism), an inference parallelization strategy that preserves data-parallel execution while offloading MoE weights across peer GPUs and fetching missing experts on demand. By removing collective inter-rank synchronization, DWDP allows each GPU to progress independently. We further address the practical overheads of this design with two optimizations for split-weight management and asynchronous remote-weight prefetch. Implemented in TensorRT-LLM and evaluated with DeepSeek-R1 on GB200 NVL72, DWDP improves end-to-end output TPS/GPU by 8.8% at comparable TPS/user in the 20-100 TPS/user serving range under 8K input sequence length and 1K output sequence length.

cs.AI, q-bio.NC updates on arXiv.org
Robust Embodied Perception in Dynamic Environments via Disentangled Weight Fusion Juncen Guo · Xiaoguang Zhu · Jingyi Wu · Jingyu Zhang · Jingnan Cai · Zhenghao Niu · Liang Song
arXiv:2604.01669v1 Announce Type: cross Abstract: Embodied perception systems face severe challenges of dynamic environment distribution drift when they continuously interact in open physical spaces. However, the existing domain incremental awareness methods often rely on the domain id obtained in advance during the testing phase, which limits their practicability in unknown interaction scenarios. At the same time, the model often overfits to the context-specific perceptual noise, which leads t
3 April 2026 at 12:00

Robust Embodied Perception in Dynamic Environments via Disentangled Weight Fusion

cs.AI, q-bio.NC updates on arXiv.org

By: Juncen Guo · Xiaoguang Zhu · Jingyi Wu · Jingyu Zhang · Jingnan Cai · Zhenghao Niu · Liang Song

3 April 2026 at 12:00

arXiv:2604.01669v1 Announce Type: cross Abstract: Embodied perception systems face severe challenges of dynamic environment distribution drift when they continuously interact in open physical spaces. However, the existing domain incremental awareness methods often rely on the domain id obtained in advance during the testing phase, which limits their practicability in unknown interaction scenarios. At the same time, the model often overfits to the context-specific perceptual noise, which leads to insufficient generalization ability and catastrophic forgetting. To address these limitations, we propose a domain-id and exemplar-free incremental learning framework for embodied multimedia systems, which aims to achieve robust continuous environment adaptation. This method designs a disentangled representation mechanism to remove non-essential environmental style interference, and guide the model to focus on extracting semantic intrinsic features shared across scenes, thereby eliminating perceptual uncertainty and improving generalization. We further use the weight fusion strategy to dynamically integrate the old and new environment knowledge in the parameter space, so as to ensure that the model adapts to the new distribution without storing historical data and maximally retains the discrimination ability of the old environment. Extensive experiments on multiple standard benchmark datasets show that the proposed method significantly reduces catastrophic forgetting in a completely exemplar-free and domain-id free setting, and its accuracy is better than the existing state-of-the-art methods.

Normal view