❌

Normal view

Received — 15 September 2026 ⏭

SimSkill: A Self-Evolving LLM Agent for Skill and Knowledge Accumulation in Traffic Simulation

✇cs.AI, q-bio.NC updates on arXiv.org
By: Qi Liu · Qinzheng Wang · Can Li · Yiming Bie · Wanjing Ma
14 September 2026 at 12:00
arXiv:2609.03753v3 Announce Type: replace Abstract: Cumulative culture enables humans to preserve, reuse, and extend knowledge and skills across experiences and generations. Inspired by this principle, we introduce \textit{SimSkill}, a self-evolving agent built around the Simulation of Urban MObility (SUMO) traffic simulator. SimSkill continually identifies capability gaps, generates and solves environment-grounded tasks, verifies solutions through an action--critic loop, and consolidates experience into episodic, procedural, and semantic memory. Through autonomous exploration, it builds a library of reusable skills and knowledge spanning major stages of the traffic-simulation workflow. We evaluate SimSkill on two held-out benchmarks across three backbone LLMs, with each result independently verified. It improves verified success by up to 25 percentage points, and ablations show complementary contributions from procedural and semantic memory. Its benefits remain backbone- and budget-dependent, as memory does not improve every model or uniformly reduce inference cost. More broadly, SimSkill illustrates a natural-language-centered design paradigm for LLM-based agent systems. Its high-level control logic, operating principles, and accumulated knowledge are expressed in natural language, while an LLM integrates them with executable tools and code to realize precise and reproducible execution. All code and experimental data are publicly available at https://github.com/qiliuchn/SimSkill-V1.
Received — 27 May 2026 ⏭

Geo-Expert: Towards Expert-Level Geological Reasoning via Parameter-Efficient Fine-Tuning

✇cs.AI, q-bio.NC updates on arXiv.org
By: Chenyou Guo · Zongqi Liu · Yizhou Zhang · Zhaorui Jiang · Ze Liu
26 May 2026 at 12:00
arXiv:2605.24844v1 Announce Type: new Abstract: While general-purpose Large Language Models (LLMs) applied to Geology often hallucinate when reasoning about subsurface structures and deep-time evolution, current AI in Earth sciences predominantly targets surface remote sensing and GIS. To bridge this gap, we introduce Geo-Expert, a family of parameter-efficient geological LLMs fine-tuned on a custom-curated, high-quality instruction dataset processed using our custom instruction synthesis pipeline. We investigate the impact of model scaling and architecture by fine-tuning three base models: Qwen3-8B, Qwen3-32B, and Gemma-3-27B, with Low-Rank Adaptation (LoRA) method. Our extensive evaluation on a novel domain-specific benchmark, Geo-Eval, reveals that a domain-aligned 8B model can outperform open-weight 70B generalists and proprietary GPT-4o on specialized geological reasoning, while a 32B variant approaches frontier reasoning models. The optimized 8B model further offers a competitive cost-performance ratio for deployment. This work provides a reproducible recipe for democratizing scientific LLMs and establishes a baseline for geological artificial intelligence.

Cross-Domain Energy-Guided Diffusion Generation for Off-Dynamics Reinforcement Learning

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yu Yang · Yihong Guo · Anqi Liu · Pan Xu
26 May 2026 at 12:00
arXiv:2605.24810v1 Announce Type: cross Abstract: Off-dynamics offline reinforcement learning seeks to learn a target-domain policy from a large source dataset and a limited target dataset under mismatched transition dynamics. Existing approaches such as reward augmentation and data filtering are constrained to the source dataset and cannot synthesize new target behavior to improve coverage beyond the collected source trajectories. While recent model-based methods attempt to address this by learning target-aware dynamics, the generated experience is constructed only at the transition level, which leads to accumulated errors over long horizons. These limitations necessitate a shift toward trajectory-level generation for off-dynamics offline RL. We propose CEDGE, a Cross-domain Energy-guided Diffusion GEneration framework. CEDGE trains a trajectory diffusion model on source-domain trajectories and adapts the generated samples to the target domain through energy guidance. This guidance is derived by minimizing the distribution mismatch between the source and desired target-domain trajectories and is decomposed into return, domain, and behavior energy components. The resulting energy-guided trajectories are useful both for direct planning and as synthetic data for policy learning. Since target adaptation is achieved via energy guidance rather than retraining the diffusion model, CEDGE can be efficiently adapted to new target dynamics compared to previous methods. Experiments on the ODRL benchmark demonstrate that trajectory-level energy-guided generation improves diffusion planning under dynamics shifts and produces synthetic data that improves downstream target policy learning.

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.

STAPO: Stabilizing Reinforcement Learning for LLMs by Silencing Rare Spurious Tokens

✇cs.AI, q-bio.NC updates on arXiv.org
By: Shiqi Liu · Zeyu He · Guojian Zhan · Letian Tao · Zhilong Zheng · Jiang Wu · Yinuo Wang · Yang Guan · Kehua Sheng · Bo Zhang · Keqiang Li · Jingliang Duan · Shengbo Eben Li
26 May 2026 at 12:00
arXiv:2602.15620v5 Announce Type: replace-cross Abstract: Reinforcement Learning (RL) has significantly improved large language model reasoning, but existing RL fine-tuning methods rely heavily on heuristic techniques such as entropy regularization and reweighting to maintain stability. In practice, they often suffer from late-stage performance collapse, leading to degraded reasoning quality and unstable training. We identify a key factor behind this instability: a small fraction of tokens, termed spurious tokens (around 0.01%), which contribute little to the reasoning outcome but receive disproportionately amplified gradient updates due to inheriting the full sequence-level reward. We present a unified framework for evaluating token-level optimization impacts across spurious risk, gradient norms, and entropy changes. Building on the analysis of token characteristics that severely disrupt optimization, we propose the Silencing Spurious Tokens (S2T) mechanism to efficiently suppress their gradient perturbations. Incorporating this mechanism into a group-based objective, we propose Spurious-Token-Aware Policy Optimization (STAPO), which promotes stable and effective large-scale model refinement. Across six mathematical reasoning benchmarks using Qwen 1.7B, 8B, and 14B base models, STAPO consistently demonstrates superior entropy stability and achieves an average performance improvement of 11.49% ($\rho_{\mathrm{T}}$=1.0, top-p=1.0) and 3.73% ($\rho_{\mathrm{T}}$=0.7, top-p=0.9) over GRPO, 20-Entropy, and JustRL.

Proteomic and lipidomic analyses reveal molecular subtypes and potential targets in early-stage lung adenocarcinoma among non-smokers

✇Pulmonary nodule
By: Yaohui Sun · Jian Liu · Xin Xu · Qiqi Liu · Hua Cheng · Wenwen Huo · Ronghao Wang · Qingdong Cao
29 April 2026 at 18:00

Cell Rep. 2026 May 26;45(5):117215. doi: 10.1016/j.celrep.2026.117215. Epub 2026 Apr 28.

ABSTRACT

Early-stage lung adenocarcinoma (LUAD) in never smokers exhibits distinct biological features, yet the metabolic programs driving early invasion remain unclear. We integrate proteomic and lipidomic profiling of primary LUAD tumors from never smokers, matched normal adjacent tissues (NATs), and benign pulmonary nodules (BPNs). Integrated multi-omics analysis reveals coordinated dysregulation of lipid metabolism and immune signaling in early LUAD. Proteome-based network fusion stratifies invasive LUAD into immune-metabolic synergistic (IMS) and metabolic-stress-driven (MSD) subtypes. IMS tumors retain apolipoprotein-associated lipid modules and favorable immune features, whereas MSD tumors exhibit stress-response programs. Mechanistically, APOA1 and APOC1 emerge as key nodes linking lipid homeostasis to invasion, and their depletion promotes LUAD cell migration and invasion. We establish a two-protein, four-lipid diagnostic panel demonstrating robust performance across tissue and plasma cohorts. These findings provide a molecular basis for early detection and risk stratification in never smokers.

PMID:42054209 | DOI:10.1016/j.celrep.2026.117215

Received — 13 April 2026 ⏭

Establishment and characterization of an immortalized porcine gastric epithelial cell line and identification of NPC1 as a key mediator of aflatoxin B1 toxicity

✇(Multiomics OR Omics) AND (Lung OR gastric OR Hepatocellular)
By: Liqi Liu · JianLin Yuan · Jinyi Bai · Yuhan Ma · Kaitao Kong · Xiaofei He · Shuangshuang Chen · Jin Chai · Keren Long · Long Jin · Mingzhou Li · Juan Deng
12 April 2026 at 18:00

Gene. 2026 Apr 9:150160. doi: 10.1016/j.gene.2026.150160. Online ahead of print.

ABSTRACT

Porcine gastric epithelial cells (PGECs) serve as a valuable model for studying the molecular and pathogenic mechanisms of the stomach. However, PGECs face limitations such as isolation challenges, short lifespan, and restricted proliferation. To address this, we established an immortalized PGECs (i-PGECs) to enable in vitro investigation of pathogen infection mechanisms. Primary PGECs were isolated from the acid-secreting glands using stepwise digestion with multiple enzymes (dispase II/collagenase I/hyaluronidase). Immortalization was achieved via lentiviral vectors expressing simian virus 40 large T antigen (SV40T) and human telomerase reverse transcriptase (hTERT), with successful expression confirmed by qRT-PCR (P < 0.05). Epithelial identity of i-PGECs was confirmed by stable expression of CK18, EpCAM, and E-cadherin, as shown by qRT-PCR and immunofluorescence. i-PGECs retained the morphological and ultrastructural features of PGECs and exhibited enhanced proliferation, as demonstrated by WST-8 assays, apoptosis and cell cycle analysis, karyotyping, and transmission electron microscopy (TEM). Telomere length analysis and scratch wound assays demonstrated stable telomere maintenance and consistent migration capacity unaffected by passaging. RNA-sequencing and differential expressed genes (DEGs) analysis revealed significantly upregulating of genes involved in cell proliferation pathways (P < 0.01). Following aflatoxin B1 (AFB1) exposure, i-PGECs significantly upregulated immune-related factors, such as NPC1 and PLAUR (P < 0.01). CRISPR/Cas9-mediated knockout of NPC1 in i-PGECs conferred increased resistance to AFB1-induced cytotoxicity, as shown by WST-8 assay. The i-PGECs remained stable after more than 50 passages, supporting their use as a reliable for in vitro model investigating the mechanisms of toxicity infection in the porcine gastric epithelium.

PMID:41966285 | DOI:10.1016/j.gene.2026.150160

Received — 8 April 2026 ⏭

ClawArena: Benchmarking AI Agents in Evolving Information Environments

✇cs.AI, q-bio.NC updates on arXiv.org
By: Haonian Ji · Kaiwen Xiong · Siwei Han · Peng Xia · Shi Qiu · Yiyang Zhou · Jiaqi Liu · Jinlong Li · Bingzhou Li · Zeyu Zheng · Cihang Xie · Huaxiu Yao
7 April 2026 at 12:00
arXiv:2604.04202v1 Announce Type: cross Abstract: AI agents deployed as persistent assistants must maintain correct beliefs as their information environment evolves. In practice, evidence is scattered across heterogeneous sources that often contradict one another, new information can invalidate earlier conclusions, and user preferences surface through corrections rather than explicit instructions. Existing benchmarks largely assume static, single-authority settings and do not evaluate whether agents can keep up with this complexity. We introduce ClawArena, a benchmark for evaluating AI agents in evolving information environments. Each scenario maintains a complete hidden ground truth while exposing the agent only to noisy, partial, and sometimes contradictory traces across multi-channel sessions, workspace files, and staged updates. Evaluation is organized around three coupled challenges: multi-source conflict reasoning, dynamic belief revision, and implicit personalization, whose interactions yield a 14-category question taxonomy. Two question formats, multi-choice (set-selection) and shell-based executable checks, test both reasoning and workspace grounding. The current release contains 64 scenarios across 8 professional domains, totaling 1{,}879 evaluation rounds and 365 dynamic updates. Experiments on five agent frameworks and five language models show that both model capability (15.4% range) and framework design (9.2%) substantially affect performance, that self-evolving skill frameworks can partially close model-capability gaps, and that belief revision difficulty is determined by update design strategy rather than the mere presence of updates. Code is available at https://github.com/aiming-lab/ClawArena.

Representation learning to advance multi-institutional studies with electronic health record data from US and France

✇cs.AI, q-bio.NC updates on arXiv.org
By: Doudou Zhou · Han Tong · Linshanshan Wang · Suqi Liu · Xin Xiong · Ziming Gan · Romain Griffier · Boris Hejblum · Yun-Chung Liu · Chuan Hong · Clara-Lea Bonzel · Tianrun Cai · Kevin Pan · Yuk-Lam Ho · Lauren Costa · Vidul A. Panickan · J. Michael Gaziano · Kenneth Mandl · Vianney Jouhet · Rodolphe Thiebaut · Zongqi Xia · Kelly Cho · Katherine Liao · Tianxi Cai
7 April 2026 at 12:00
arXiv:2502.08547v2 Announce Type: replace Abstract: The widespread adoption of electronic health records has created new opportunities for translational clinical research, yet this promise remains constrained by fragmented data across privacy-siloed institutions and substantial heterogeneity in local coding practices. While privacy-preserving collaborative learning allows institutions to work together without sharing patient-level data, it does not address inconsistencies in how clinical concepts are represented across sites. We introduce a graph-based framework that addresses this gap by treating data harmonization as a scalable representation learning problem. Rather than relying on fixed standards or manual mappings, the framework integrates institution-specific summary statistics from health records, curated biomedical knowledge graphs, and semantic information derived from large language models to learn a shared semantic space. This joint learning approach aligns diverse, site-specific vocabularies while preserving patient privacy. Evaluated across seven institutions and two languages, the framework provides a robust, data-centric foundation for training and deploying clinical models across heterogeneous healthcare systems.
Received — 4 April 2026 ⏭

Omni-SimpleMem: Autoresearch-Guided Discovery of Lifelong Multimodal Agent Memory

✇cs.AI, q-bio.NC updates on arXiv.org
By: Jiaqi Liu · Zipeng Ling · Shi Qiu · Yanqing Liu · Siwei Han · Peng Xia · Haoqin Tu · Zeyu Zheng · Cihang Xie · Charles Fleming · Mingyu Ding · Huaxiu Yao
3 April 2026 at 12:00
arXiv:2604.01007v2 Announce Type: replace Abstract: AI agents increasingly operate over extended time horizons, yet their ability to retain, organize, and recall multimodal experiences remains a critical bottleneck. Building effective lifelong memory requires navigating a vast design space spanning architecture, retrieval strategies, prompt engineering, and data pipelines; this space is too large and interconnected for manual exploration or traditional AutoML to explore effectively. We deploy an autonomous research pipeline to discover Omni-SimpleMem, a unified multimodal memory framework for lifelong AI agents. Starting from a na\"ive baseline (F1=0.117 on LoCoMo), the pipeline autonomously executes ${\sim}50$ experiments across two benchmarks, diagnosing failure modes, proposing architectural modifications, and repairing data pipeline bugs, all without human intervention in the inner loop. The resulting system achieves state-of-the-art on both benchmarks, improving F1 by +411% on LoCoMo (0.117$\to$0.598) and +214% on Mem-Gallery (0.254$\to$0.797) relative to the initial configurations. Critically, the most impactful discoveries are not hyperparameter adjustments: bug fixes (+175%), architectural changes (+44%), and prompt engineering (+188% on specific categories) each individually exceed the cumulative contribution of all hyperparameter tuning, demonstrating capabilities fundamentally beyond the reach of traditional AutoML. We provide a taxonomy of six discovery types and identify four properties that make multimodal memory particularly suited for autoresearch, offering guidance for applying autonomous research pipelines to other AI system domains. Code is available at this https://github.com/aiming-lab/SimpleMem.
Received — 26 March 2026 ⏭
Received — 17 March 2026 ⏭

MalURLBench: A Benchmark Evaluating Agents' Vulnerabilities When Processing Web URLs

✇cs.AI, q-bio.NC updates on arXiv.org
By: Dezhang Kong · Zhuxi Wu · Shiqi Liu · Zhicheng Tan · Kuichen Lu · Minghao Li · Qichen Liu · Shengyu Chu · Zhenhua Xu · Xuan Liu · Meng Han
16 March 2026 at 12:00
arXiv:2601.18113v3 Announce Type: replace-cross Abstract: LLM-based web agents have become increasingly popular for their utility in daily life and work. However, they exhibit critical vulnerabilities when processing malicious URLs: accepting a disguised malicious URL enables subsequent access to unsafe webpages, which can cause severe damage to service providers and users. Despite this risk, no benchmark currently targets this emerging threat. To address this gap, we propose MalURLBench, the first benchmark for evaluating LLMs' vulnerabilities to malicious URLs. MalURLBench contains 61,845 attack instances spanning 10 real-world scenarios and 7 categories of real malicious websites. Experiments with 12 popular LLMs reveal that existing models struggle to detect elaborately disguised malicious URLs. We further identify and analyze key factors that impact attack success rates and propose URLGuard, a lightweight defense module. We believe this work will provide a foundational resource for advancing the security of web agents. Our code is available at https://github.com/JiangYingEr/MalURLBench.

VideoTemp-o3: Harmonizing Temporal Grounding and Video Understanding in Agentic Thinking-with-Videos

✇cs.AI, q-bio.NC updates on arXiv.org
By: Wenqi Liu · Yunxiao Wang · Shijie Ma · Meng Liu · Qile Su · Tianke Zhang · Haonan Fan · Changyi Liu · Kaiyu Jiang · Jiankang Chen · Kaiyu Tang · Bin Wen · Fan Yang · Tingting Gao · Han Li · Yinwei Wei · Xuemeng Song
16 March 2026 at 12:00
arXiv:2602.07801v3 Announce Type: replace-cross Abstract: In long-video understanding, conventional uniform frame sampling often fails to capture key visual evidence, leading to degraded performance and increased hallucinations. To address this, recent agentic thinking-with-videos paradigms have emerged, adopting a localize-clip-answer pipeline in which the model actively identifies relevant video segments, performs dense sampling within those clips, and then produces answers. However, existing methods remain inefficient, suffer from weak localization, and adhere to rigid workflows. To solve these issues, we propose VideoTemp-o3, a unified agentic thinking-with-videos framework that jointly models video grounding and question answering. VideoTemp-o3 exhibits strong localization capability, supports on-demand clipping, and can refine inaccurate localizations. Specifically, in the supervised fine-tuning stage, we design a unified masking mechanism that encourages exploration while preventing noise. For reinforcement learning, we introduce dedicated rewards to mitigate reward hacking. Besides, from the data perspective, we develop an effective pipeline to construct high-quality long video grounded QA data, along with a corresponding benchmark for systematic evaluation across various video durations. Experimental results demonstrate that our method achieves remarkable performance on both long video understanding and grounding.
Received — 11 March 2026 ⏭

Property-driven Protein Inverse Folding With Multi-Objective Preference Alignment

✇cs.AI, q-bio.NC updates on arXiv.org
By: Xiaoyang Hou · Junqi Liu · Chence Shi · Xin Liu · Zhi Yang · Jian Tang
10 March 2026 at 12:00
arXiv:2603.06748v1 Announce Type: cross Abstract: Protein sequence design must balance designability, defined as the ability to recover a target backbone, with multiple, often competing, developability properties such as solubility, thermostability, and expression. Existing approaches address these properties through post hoc mutation, inference-time biasing, or retraining on property-specific subsets, yet they are target dependent and demand substantial domain expertise or careful hyperparameter tuning. In this paper, we introduce ProtAlign, a multi-objective preference alignment framework that fine-tunes pretrained inverse folding models to satisfy diverse developability objectives while preserving structural fidelity. ProtAlign employs a semi-online Direct Preference Optimization strategy with a flexible preference margin to mitigate conflicts among competing objectives and constructs preference pairs using in silico property predictors. Applied to the widely used ProteinMPNN backbone, the resulting model MoMPNN enhances developability without compromising designability across tasks including sequence design for CATH 4.3 crystal structures, de novo generated backbones, and real-world binder design scenarios, making it an appealing framework for practical protein sequence design.

Kinematics-Aware Latent World Models for Data-Efficient Autonomous Driving

✇cs.AI, q-bio.NC updates on arXiv.org
By: Jiazhuo Li · Linjiang Cao · Qi Liu · Xi Xiong
10 March 2026 at 12:00
arXiv:2603.07264v1 Announce Type: cross Abstract: Data-efficient learning remains a central challenge in autonomous driving due to the high cost and safety risks of large-scale real-world interaction. Although world-model-based reinforcement learning enables policy optimization through latent imagination, existing approaches often lack explicit mechanisms to encode spatial and kinematic structure essential for driving tasks. In this work, we build upon the Recurrent State-Space Model (RSSM) and propose a kinematics-aware latent world model framework for autonomous driving. Vehicle kinematic information is incorporated into the observation encoder to ground latent transitions in physically meaningful motion dynamics, while geometry-aware supervision regularizes the RSSM latent state to capture task-relevant spatial structure beyond pixel reconstruction. The resulting structured latent dynamics improve long-horizon imagination fidelity and stabilize policy optimization. Experiments in a driving simulation benchmark demonstrate consistent gains over both model-free and pixel-based world-model baselines in terms of sample efficiency and driving performance. Ablation studies further verify that the proposed design enhances spatial representation quality within the latent space. These results suggest that integrating kinematic grounding into RSSM-based world models provides a scalable and physically grounded paradigm for autonomous driving policy learning.

Survey of Computerized Adaptive Testing: A Machine Learning Perspective

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yan Zhuang · Qi Liu · Haoyang Bi · Zhenya Huang · Weizhe Huang · Jiatong Li · Junhao Yu · Zirui Liu · Zirui Hu · Yuting Hong · Zachary A. Pardos · Haiping Ma · Mengxiao Zhu · Shijin Wang · Enhong Chen
10 March 2026 at 12:00
arXiv:2404.00712v3 Announce Type: replace-cross Abstract: Computerized Adaptive Testing (CAT) offers an efficient and personalized method for assessing examinee proficiency by dynamically adjusting test questions based on individual performance. Compared to traditional, non-personalized testing methods, CAT requires fewer questions and provides more accurate assessments. As a result, CAT has been widely adopted across various fields, including education, healthcare, sports, sociology, and the evaluation of AI models. While traditional methods rely on psychometrics and statistics, the increasing complexity of large-scale testing has spurred the integration of machine learning techniques. This paper aims to provide a machine learning-focused survey on CAT, presenting a fresh perspective on this adaptive testing paradigm. We delve into measurement models, question selection algorithm, bank construction, and test control within CAT, exploring how machine learning can optimize these components. Through an analysis of current methods, strengths, limitations, and challenges, we strive to develop robust, fair, and efficient CAT systems. By bridging psychometric-driven CAT research with machine learning, this survey advocates for a more inclusive and interdisciplinary approach to the future of adaptive testing.
Received — 5 March 2026 ⏭

Neural Paging: Learning Context Management Policies for Turing-Complete Agents

✇cs.AI, q-bio.NC updates on arXiv.org
By: Liang Chen · Qi Liu
4 March 2026 at 13:00
arXiv:2603.02228v1 Announce Type: cross Abstract: The proof that Large Language Models (LLMs) augmented with external read-write memory constitute a computationally universal system has established the theoretical foundation for general-purpose agents. However, existing implementations face a critical bottleneck: the finite and costly Context Window, which functions not as infinite memory but as a scarce semantic cache. In this work, we introduce \textit{Neural Paging}, a hierarchical architecture that decouples symbolic reasoning from information resource management. We formulate the \textit{Context Paging Problem (CPP)} and propose a lightweight, differentiable \textit{Page Controller} designed to approximate ``Semantic Belady's Optimality'' -- retaining tokens with high future utility under explicit assumptions on access patterns. We provide theoretical analysis showing that, under bounded context window size~$K$, Neural Paging reduces the asymptotic complexity of long-horizon reasoning from quadratic $O(N^2)$ to $O(N \cdot K^2)$, and we derive a robustness bound (Theorem~4) that quantifies competitive-ratio degradation under policy-dependent access with bounded sensitivity. We validate these bounds on synthetic paging traces, confirming that the theoretical guarantees hold and identifying significant slack that motivates learned policies.

Perception-R1: Advancing Multimodal Reasoning Capabilities of MLLMs via Visual Perception Reward

✇cs.AI, q-bio.NC updates on arXiv.org
By: Tong Xiao · Xin Xu · Zhenya Huang · Hongyu Gao · Quan Liu · Qi Liu · Enhong Chen
4 March 2026 at 13:00
arXiv:2506.07218v3 Announce Type: replace-cross Abstract: Enhancing the multimodal reasoning capabilities of Multimodal Large Language Models (MLLMs) is a challenging task that has attracted increasing attention in the community. Recently, several studies have applied Reinforcement Learning with Verifiable Rewards (RLVR) to the multimodal domain in order to enhance the reasoning abilities of MLLMs. However, these works largely overlook the enhancement of multimodal perception capabilities in MLLMs, which serve as a core prerequisite and foundational component of complex multimodal reasoning. Through McNemar's test, we find that existing RLVR method fails to effectively enhance the multimodal perception capabilities of MLLMs, thereby limiting their further improvement in multimodal reasoning. To address this limitation, we propose Perception-R1, which introduces a novel visual perception reward that explicitly encourages MLLMs to perceive the visual content accurately, thereby can effectively incentivizing both their multimodal perception and reasoning capabilities. Specifically, we first collect textual visual annotations from the CoT trajectories of multimodal problems, which will serve as visual references for reward assignment. During RLVR training, we employ a judging LLM to assess the consistency between the visual annotations and the responses generated by MLLM, and assign the visual perception reward based on these consistency judgments. Extensive experiments on several multimodal reasoning benchmarks demonstrate the effectiveness of our Perception-R1, which achieves state-of-the-art performance on most benchmarks using only 1,442 training data. Our code and dataset will be available at https://github.com/tongxiao2002/Perception-R1.

VideoTemp-o3: Harmonizing Temporal Grounding and Video Understanding in Agentic Thinking-with-Videos

✇cs.AI, q-bio.NC updates on arXiv.org
By: Wenqi Liu · Yunxiao Wang · Shijie Ma · Meng Liu · Qile Su · Tianke Zhang · Haonan Fan · Changyi Liu · Kaiyu Jiang · Jiankang Chen · Kaiyu Tang · Bin Wen · Fan Yang · Tingting Gao · Han Li · Yinwei Wei · Xuemeng Song
4 March 2026 at 13:00
arXiv:2602.07801v2 Announce Type: replace-cross Abstract: In long-video understanding, conventional uniform frame sampling often fails to capture key visual evidence, leading to degraded performance and increased hallucinations. To address this, recent agentic thinking-with-videos paradigms have emerged, adopting a localize-clip-answer pipeline in which the model actively identifies relevant video segments, performs dense sampling within those clips, and then produces answers. However, existing methods remain inefficient, suffer from weak localization, and adhere to rigid workflows. To solve these issues, we propose VideoTemp-o3, a unified agentic thinking-with-videos framework that jointly models video grounding and question answering. VideoTemp-o3 exhibits strong localization capability, supports on-demand clipping, and can refine inaccurate localizations. Specifically, in the supervised fine-tuning stage, we design a unified masking mechanism that encourages exploration while preventing noise. For reinforcement learning, we introduce dedicated rewards to mitigate reward hacking. Besides, from the data perspective, we develop an effective pipeline to construct high-quality long video grounded QA data, along with a corresponding benchmark for systematic evaluation across various video durations. Experimental results demonstrate that our method achieves remarkable performance on both long video understanding and grounding.
Received — 25 February 2026 ⏭

Physiologically Informed Deep Learning: A Multi-Scale Framework for Next-Generation PBPK Modeling

✇cs.AI, q-bio.NC updates on arXiv.org
By: Shunqi Liu · Han Qiu · Tong Wang
24 February 2026 at 13:00
arXiv:2602.18472v1 Announce Type: cross Abstract: Physiologically Based Pharmacokinetic (PBPK) modeling is a cornerstone of model-informed drug development (MIDD), providing a mechanistic framework to predict drug absorption, distribution, metabolism, and excretion (ADME). Despite its utility, adoption is hindered by high computational costs for large-scale simulations, difficulty in parameter identification for complex biological systems, and uncertainty in interspecies extrapolation. In this work, we propose a unified Scientific Machine Learning (SciML) framework that bridges mechanistic rigor and data-driven flexibility. We introduce three contributions: (1) Foundation PBPK Transformers, which treat pharmacokinetic forecasting as a sequence modeling task; (2) Physiologically Constrained Diffusion Models (PCDM), a generative approach that uses a physics-informed loss to synthesize biologically compliant virtual patient populations; and (3) Neural Allometry, a hybrid architecture combining Graph Neural Networks (GNNs) with Neural ODEs to learn continuous cross-species scaling laws. Experiments on synthetic datasets show that the framework reduces physiological violation rates from 2.00% to 0.50% under constraints while offering a path to faster simulation.
❌