❌

Normal view

Received — 27 May 2026 ⏭

Dynamic Dual-Granularity Skill Bank for Agentic RL

✇cs.AI, q-bio.NC updates on arXiv.org
By: Songjun Tu · Chengdong Xu · Qichao Zhang · Yaocheng Zhang · Xiangyuan Lan · Linjing Li · Dong Li · Dongbin Zhao
26 May 2026 at 12:00
arXiv:2603.28716v2 Announce Type: replace Abstract: Agentic RL can benefit substantially from reusable experience, yet existing skill-based methods mainly extract trajectory-level guidance and often lack principled mechanisms for maintaining an evolving skill memory. We propose D2Skill, a dynamic dual-granularity skill bank for agentic RL that organizes reusable experience into task skills for high-level guidance and step skills for fine-grained decision support and error correction. D2Skill jointly trains the policy and skill bank through paired baseline and skill-injected rollouts under the same policy, using their performance gap to derive hindsight utility signals for both skill updating and policy optimization. Built entirely from training-time experience, the skill bank is continuously expanded through reflection and maintained with utility-aware retrieval and pruning. Experiments on ALFWorld, WebShop, and Search-Augmented QA tasks show that D2Skill substantially improves performance over skill-free baselines across models of different scales. Further ablations and analyses show that both dual-granularity skill modeling and dynamic skill maintenance are critical to these gains, while the learned skills exhibit higher utility, transfer across evaluation settings, and introduce only modest training overhead.

SEA-Eval: A Benchmark for Evaluating Self-Evolving Agents Beyond Episodic Assessment

✇cs.AI, q-bio.NC updates on arXiv.org
By: Sihang Jiang · Lipeng Ma · Zhonghua Hong · Keyi Wang · Zhiyu Lu · Tengfei Wang · Shisong Chen · Jinghao Zhang · Tianjun Pan · Weijia Li · Jiaqing Liang · Yanghua Xiao
26 May 2026 at 12:00
arXiv:2604.08988v3 Announce Type: replace Abstract: Current LLM-based agents demonstrate strong performance in episodic task execution but remain constrained by static toolsets and episodic amnesia, failing to accumulate experience across task boundaries. This paper formalizes the Self-Evolving Agent (SEA) from the perspective of digital embodiment and continuous cross-task evolution, introduces the Evolutionary Flywheel as its minimal sufficient architecture, and presents SEA-Eval -- the first benchmark designed specifically for evaluating SEAs. Grounded in Flywheel theory, SEA-Eval establishes SR and T as primary metrics and, through sequential task stream design, is designed to quantify evolutionary gain, evolutionary stability, and implicit alignment convergence. Empirical evaluation reveals that, under comparable success rates, token consumption differs by up to 31.2 times between frameworks on individual tasks, with divergent evolutionary trajectories emerging under sequential analysis -- demonstrating that success rate alone creates a capability illusion and that the sequential convergence of $T$ is the key criterion for distinguishing genuine evolution from pseudo-evolution.

JAEGER: Joint 3D Audio-Visual Grounding and Reasoning in Simulated Physical Environments

✇cs.AI, q-bio.NC updates on arXiv.org
By: Zhan Liu · Changli Tang · Yuxin Wang · Zhiyuan Zhu · Youjun Chen · Yiwen Shao · Tianzi Wang · Lei Ke · Zengrui Jin · Chao Zhang
26 May 2026 at 12:00
arXiv:2602.18527v2 Announce Type: replace-cross Abstract: Current audio-visual large language models (AV-LLMs) are predominantly restricted to 2D perception, relying on RGB video and monaural audio. This design choice introduces a fundamental dimensionality mismatch that precludes reliable source localization and spatial reasoning in complex 3D environments. We address this limitation by presenting JAEGER, a framework that extends AV-LLMs to 3D space, to enable joint spatial grounding and reasoning through the integration of RGB-D observations and multi-channel first-order ambisonics. A core contribution of our work is the neural intensity vector (Neural IV), a learned spatial audio representation that encodes robust directional cues to enhance direction-of-arrival estimation, even in adverse acoustic scenarios with overlapping sources. To facilitate large-scale training and systematic evaluation, we propose SpatialSceneQA, a benchmark of 61k instruction-tuning samples curated from simulated physical environments. Extensive experiments demonstrate that our approach consistently surpasses 2D-centric baselines across diverse spatial perception and reasoning tasks, underscoring the necessity of explicit 3D modelling for advancing AI in physical environments. Our source code, pre-trained model checkpoints, and datasets are available at https://github.com/liuzhan22/JAEGER.

Data Difficulty and the Generalization--Extrapolation Tradeoff in LLM Fine-Tuning

✇cs.AI, q-bio.NC updates on arXiv.org
By: Siyuan Liu · Tinghong Chen · Xinghan Li · Yifei Wang · Jingzhao Zhang
26 May 2026 at 12:00
arXiv:2605.12906v2 Announce Type: replace-cross Abstract: Data selection during supervised fine-tuning (SFT) can critically change the behavior of large language models (LLMs). Although existing work has studied the effect of selecting data based on heuristics such as perplexity, difficulty, or length, the reported findings are often inconsistent or context-dependent. In this work, we systematically study the role of data difficulty in fine-tuning from both empirical and theoretical perspectives, and find that there is no universally optimal difficulty level; rather, its effectiveness depends on the dataset size. We show that for a fixed data budget, there exists an optimal data difficulty for SFT, and that this optimal difficulty shifts toward harder data as the data budget increases. To explain this phenomenon, we conduct controlled synthetic experiments that reveal a simple underlying mechanism: the interplay between the (in-distribution) generalization gap and the extrapolation gap. We further support this mechanism through a theoretical analysis using PAC-Bayesian generalization bounds. Overall, our results clarify how data size and difficulty jointly affect the trade-off between generalization and extrapolation in SFT, providing guidance for difficulty-based data selection under certain model and data conditions.

FDX1 as a predictive biomarker and therapeutic target for lymph node metastasis in gastric cancer

✇Omics in Gastric
By: Tianhao Zhang · Jianqiu Liu · Peng Duan · Zikang Li · Zhixin Huang · Zeyu Zhao · Yang Cheng · Zhi Liang · Jianhui Chen · Meilan Cai · Jinping Ma
10 May 2026 at 18:00

Clin Exp Med. 2026 May 10. doi: 10.1007/s10238-026-02160-0. Online ahead of print.

ABSTRACT

The prognostic values of cuproptosis-related genes (CRGs) in gastric cancer with lymph node metastasis (GCLM), especially in the tumor immune microenvironment (TIME), remain unclear. We analyzed the expression, mutation, immunity, drug sensitivity, and prognostic value of CRGs in GCLM using TCGA and GEO cohorts. Consensus clustering was performed to identify CRG subtypes, with differences characterized by multi-omics analysis. A CRG-based prognostic risk score and immune score were constructed for individualized assessment, and the role of CRGs was validated through in vitro and in vivo experiments. Consensus clustering revealed that CRGs were significantly enriched in biological processes related to mitosis and energy metabolism, as well as in immune-related and cancer-associated pathways. Four distinct CRG subtypes were identified, showing marked differences in expression profiles, prognosis, genetic alterations, TIME, and chemotherapeutic drug sensitivity. We developed an exploratory CRG-based prognostic risk score for preliminary individualized assessment, and the functional relevance of CRGs in GCLM was further validated through in vitro experiments. Among these, FDX1, LIAS, DLAT, MTF1, and GLS were identified as key determinants of overall survival in patients with GCLM, with FDX1 emerging as a potential independent prognostic factor. Notably, upregulation of FDX1 significantly suppressed lymph node metastasis of gastric cancer cells in a mouse popliteal lymph node metastasis model. Our data uncovers FDX1 might be a potential favorable prognostic factors in GCLM patients. These findings may improve our understanding of CRGs in GCLM and provide new in-sights for assessing prognosis and developing more effective treatment strategies.

PMID:42107026 | DOI:10.1007/s10238-026-02160-0

Received — 8 April 2026 ⏭

ActionNex: A Virtual Outage Manager for Cloud

✇cs.AI, q-bio.NC updates on arXiv.org
By: Zhenfeng Lin · Haoji Hu · Ming Hao · Xuchao Zhang · Ryan Zhang · Junhao Li · Ze Li · Oleg Kulygin · Chetan Bansal · Hatay Tuna · Murali Chintalapati · Sheila Jiang · Salman Zafar · Angie Anderson
7 April 2026 at 12:00
arXiv:2604.03512v1 Announce Type: new Abstract: Outage management in large-scale cloud operations remains heavily manual, requiring rapid triage, cross-team coordination, and experience-driven decisions under partial observability. We present \textbf{ActionNex}, a production-grade agentic system that supports end-to-end outage assistance, including real-time updates, knowledge distillation, and role- and stage-conditioned next-best action recommendations. ActionNex ingests multimodal operational signals (e.g., outage content, telemetry, and human communications) and compresses them into critical events that represent meaningful state transitions. It couples this perception layer with a hierarchical memory subsystem: long-term Key-Condition-Action (KCA) knowledge distilled from playbooks and historical executions, episodic memory of prior outages, and working memory of the live context. A reasoning agent aligns current critical events to preconditions, retrieves relevant memories, and generates actionable recommendations; executed human actions serve as an implicit feedback signal to enable continual self-evolution in a human-agent hybrid system. We evaluate ActionNex on eight real Azure outages (8M tokens, 4,000 critical events) using two complementary ground-truth action sets, achieving 71.4\% precision and 52.8-54.8\% recall. The system has been piloted in production and has received positive early feedback.

Schema-Aware Planning and Hybrid Knowledge Toolset for Reliable Knowledge Graph Triple Verification

✇cs.AI, q-bio.NC updates on arXiv.org
By: Xinyan Ma (Harbin Institute of Technology · Harbin · China) · Xianhao Ou (Harbin Institute of Technology · Harbin · China) · Weihao Zhang (Harbin Institute of Technology · Harbin · China) · Shixin Jiang (Harbin Institute of Technology · Harbin · China) · Runxuan Liu (Harbin Institute of Technology · Harbin · China) · Dandan Tu (Huawei Technologies Co. · Ltd. · Beijing · China) · Lei Chen (Bay Area International Business School · Beijing Normal University · Beijing · China) · Ming Liu (Harbin Institute of Technology · Harbin · China) · Bing Qin (Harbin Institute of Technology · Harbin · China)
7 April 2026 at 12:00
arXiv:2604.04190v1 Announce Type: new Abstract: Knowledge Graphs (KGs) serve as a critical foundation for AI systems, yet their automated construction inevitably introduces noise, compromising data trustworthiness. Existing triple verification methods, based on graph embeddings or language models, often suffer from single-source bias by relying on either internal structural constraints or external semantic evidence, and usually follow a static inference paradigm. As a result, they struggle with complex or long-tail facts and provide limited interpretability. To address these limitations, we propose SHARP (Schema-Hybrid Agent for Reliable Prediction), a training-free autonomous agent that reformulates triple verification as a dynamic process of strategic planning, active investigation, and evidential reasoning. Specifically, SHARP combines a Memory-Augmented Mechanism with Schema-Aware Strategic Planning to improve reasoning stability, and employs an enhanced ReAct loop with a Hybrid Knowledge Toolset to dynamically integrate internal KG structure and external textual evidence for cross-verification. Experiments on FB15K-237 and Wikidata5M-Ind show that SHARP significantly outperforms existing state-of-the-art baselines, achieving accuracy gains of 4.2% and 12.9%, respectively. Moreover, SHARP provides transparent, fact-based evidence chains for each judgment, demonstrating strong interpretability and robustness for complex verification tasks.

Diagonal-Tiled Mixed-Precision Attention for Efficient Low-Bit MXFP Inference

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yifu Ding · Xinhao Zhang · Jinyang Guo
7 April 2026 at 12:00
arXiv:2604.03950v1 Announce Type: cross Abstract: Transformer-based large language models (LLMs) have demonstrated remarkable performance across a wide range of real-world tasks, but their inference cost remains prohibitively high due to the quadratic complexity of attention and the memory bandwidth limitations of high-precision operations. In this work, we present a low-bit mixed-precision attention kernel using the microscaling floating-point (MXFP) data format, utilizing the computing capability on next-generation GPU architectures. Our Diagonal-Tiled Mixed-Precision Attention (DMA) incorporates two kinds of low-bit computation at the tiling-level, and is a delicate fused kernel implemented using Triton, exploiting hardware-level parallelism and memory efficiency to enable fast and efficient inference without compromising model performance. Extensive empirical evaluations on NVIDIA B200 GPUs show that our kernel maintains generation quality with negligible degradation, and meanwhile achieves significant speedup by kernel fusion. We release our code at https://github.com/yifu-ding/MP-Sparse-Attn.

ROSClaw: A Hierarchical Semantic-Physical Framework for Heterogeneous Multi-Agent Collaboration

✇cs.AI, q-bio.NC updates on arXiv.org
By: Rongfeng Zhao · Xuanhao Zhang · Zhaochen Guo · Xiang Shao · Zhongpan Zhu · Bin He · Jie Chen
7 April 2026 at 12:00
arXiv:2604.04664v1 Announce Type: cross Abstract: The integration of large language models (LLMs) with embodied agents has improved high-level reasoning capabilities; however, a critical gap remains between semantic understanding and physical execution. While vision-language-action (VLA) and vision-language-navigation (VLN) systems enable robots to perform manipulation and navigation tasks from natural language instructions, they still struggle with long-horizon sequential and temporally structured tasks. Existing frameworks typically adopt modular pipelines for data collection, skill training, and policy deployment, resulting in high costs in experimental validation and policy optimization. To address these limitations, we propose ROSClaw, an agent framework for heterogeneous robots that integrates policy learning and task execution within a unified vision-language model (VLM) controller. The framework leverages e-URDF representations of heterogeneous robots as physical constraints to construct a sim-to-real topological mapping, enabling real-time access to the physical states of both simulated and real-world agents. We further incorporate a data collection and state accumulation mechanism that stores robot states, multimodal observations, and execution trajectories during real-world execution, enabling subsequent iterative policy optimization. During deployment, a unified agent maintains semantic continuity between reasoning and execution, and dynamically assigns task-specific control to different agents, thereby improving robustness in multi-policy execution. By establishing an autonomous closed-loop framework, ROSClaw minimizes the reliance on robot-specific development workflows. The framework supports hardware-level validation, automated generation of SDK-level control programs, and tool-based execution, enabling rapid cross-platform transfer and continual improvement of robotic skills. Ours project page: https://www.rosclaw.io/.

Mind Your HEARTBEAT! Claw Background Execution Inherently Enables Silent Memory Pollution

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yechao Zhang · Shiqian Zhao · Jie Zhang · Gelei Deng · Jiawen Zhang · Xiaogeng Liu · Chaowei Xiao · Tianwei Zhang
7 April 2026 at 12:00
arXiv:2603.23064v3 Announce Type: replace-cross Abstract: We identify a critical security vulnerability in mainstream Claw personal AI agents: untrusted content encountered during heartbeat-driven background execution can silently pollute agent memory and subsequently influence user-facing behavior without the user's awareness. This vulnerability arises from an architectural design shared across the Claw ecosystem: heartbeat background execution runs in the same session as user-facing conversation, so content ingested from any external source monitored in the background (including email, message channels, news feeds, code repositories, and social platforms) can enter the same memory context used for foreground interaction, often with limited user visibility and without clear source provenance. We formalize this process as an Exposure (E) $\rightarrow$ Memory (M) $\rightarrow$ Behavior (B) pathway: misinformation encountered during heartbeat execution enters the agent's short-term session context, potentially gets written into long-term memory, and later shapes downstream user-facing behavior. We instantiate this pathway in an agent-native social setting using MissClaw, a controlled research replica of Moltbook. We find that (1) social credibility cues, especially perceived consensus, are the dominant driver of short-term behavioral influence, with misleading rates up to 61%; (2) routine memory-saving behavior can promote short-term pollution into durable long-term memory at rates up to 91%, with cross-session behavioral influence reaching 76%; (3) under naturalistic browsing with content dilution and context pruning, pollution still crosses session boundaries. Overall, prompt injection is not required: ordinary social misinformation is sufficient to silently shape agent memory and behavior under heartbeat-driven background execution.
Received — 4 April 2026 ⏭

Grounded Token Initialization for New Vocabulary in LMs for Generative Recommendation

✇cs.AI, q-bio.NC updates on arXiv.org
By: Daiwei Chen · Zhoutong Fu · Chengming Jiang · Haichao Zhang · Ran Zhou · Tan Wang · Chunnan Yao · Guoyao Li · Rui Cai · Yihan Cao · Ruijie Jiang · Fedor Borisyuk · Jianqiang Shen · Jingwei Wu · Ramya Korlakai Vinayak
3 April 2026 at 12:00
arXiv:2604.02324v1 Announce Type: cross Abstract: Language models (LMs) are increasingly extended with new learnable vocabulary tokens for domain-specific tasks, such as Semantic-ID tokens in generative recommendation. The standard practice initializes these new tokens as the mean of existing vocabulary embeddings, then relies on supervised fine-tuning to learn their representations. We present a systematic analysis of this strategy: through spectral and geometric diagnostics, we show that mean initialization collapses all new tokens into a degenerate subspace, erasing inter-token distinctions that subsequent fine-tuning struggles to fully recover. These findings suggest that \emph{token initialization} is a key bottleneck when extending LMs with new vocabularies. Motivated by this diagnosis, we propose the \emph{Grounded Token Initialization Hypothesis}: linguistically grounding novel tokens in the pretrained embedding space before fine-tuning better enables the model to leverage its general-purpose knowledge for novel-token domains. We operationalize this hypothesis as GTI (Grounded Token Initialization), a lightweight grounding stage that, prior to fine-tuning, maps new tokens to distinct, semantically meaningful locations in the pretrained embedding space using only paired linguistic supervision. Despite its simplicity, GTI outperforms both mean initialization and existing auxiliary-task adaptation methods in the majority of evaluation settings across multiple generative recommendation benchmarks, including industry-scale and public datasets. Further analyses show that grounded embeddings produce richer inter-token structure that persists through fine-tuning, corroborating the hypothesis that initialization quality is a key bottleneck in vocabulary extension.
Received — 2 April 2026 ⏭

ATP-Bench: Towards Agentic Tool Planning for MLLM Interleaved Generation

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yinuo Liu · Zi Qian · Heng Zhou · Jiahao Zhang · Yajie Zhang · Zhihang Li · Mengyu Zhou · Erchao Zhao · Xiaoxi Jiang · Guanjun Jiang
1 April 2026 at 12:00
arXiv:2603.29902v1 Announce Type: new Abstract: Interleaved text-and-image generation represents a significant frontier for Multimodal Large Language Models (MLLMs), offering a more intuitive way to convey complex information. Current paradigms rely on either image generation or retrieval augmentation, yet they typically treat the two as mutually exclusive paths, failing to unify factuality with creativity. We argue that the next milestone in this field is Agentic Tool Planning, where the model serves as a central controller that autonomously determines when, where, and which tools to invoke to produce interleaved responses for visual-critical queries. To systematically evaluate this paradigm, we introduce ATP-Bench, a novel benchmark comprising 7,702 QA pairs (including 1,592 VQA pairs) across eight categories and 25 visual-critical intents, featuring human-verified queries and ground truths. Furthermore, to evaluate agentic planning independent of end-to-end execution and changing tool backends, we propose a Multi-Agent MLLM-as-a-Judge (MAM) system. MAM evaluates tool-call precision, identifies missed opportunities for tool use, and assesses overall response quality without requiring ground-truth references. Our extensive experiments on 10 state-of-the-art MLLMs reveal that models struggle with coherent interleaved planning and exhibit significant variations in tool-use behavior, highlighting substantial room for improvement and providing actionable guidance for advancing interleaved generation. Dataset and code are available at https://github.com/Qwen-Applications/ATP-Bench.

MultiGen: Level-Design for Editable Multiplayer Worlds in Diffusion Game Engines

✇cs.AI, q-bio.NC updates on arXiv.org
By: Ryan Po · David Junhao Zhang · Amir Hertz · Gordon Wetzstein · Neal Wadhwa · Nataniel Ruiz
1 April 2026 at 12:00
arXiv:2603.06679v2 Announce Type: replace Abstract: Video world models have shown immense promise for interactive simulation and entertainment, but current systems still struggle with two important aspects of interactivity: user control over the environment for reproducible, editable experiences, and shared inference where players hold influence over a common world. To address these limitations, we introduce an explicit external memory into the system, a persistent state operating independent of the model's context window, that is continually updated by user actions and queried throughout the generation roll-out. Unlike conventional diffusion game engines that operate as next-frame predictors, our approach decomposes generation into Memory, Observation, and Dynamics modules. This design gives users direct, editable control over environment structure via an editable memory representation, and it naturally extends to real-time multiplayer rollouts with coherent viewpoints and consistent cross-player interactions.

QuestA: Expanding Reasoning Capacity in LLMs via Question Augmentation

✇cs.AI, q-bio.NC updates on arXiv.org
By: Jiazheng Li · Hongzhou Lin · Hong Lu · Kaiyue Wen · Zaiwen Yang · Jiaxuan Gao · Yi Wu · Jingzhao Zhang
1 April 2026 at 12:00
arXiv:2507.13266v4 Announce Type: replace-cross Abstract: Reinforcement learning (RL) has emerged as a central paradigm for training large language models (LLMs) in reasoning tasks. Yet recent studies question RL's ability to incentivize reasoning capacity beyond the base model. This raises a key challenge: how can RL be adapted to solve harder reasoning problems more effectively? To address this challenge, we propose a simple yet effective strategy via Question Augmentation: introduce partial solutions during training to reduce problem difficulty and provide more informative learning signals. Our method, QuestA, when applied during RL training on math reasoning tasks, not only improves pass@1 but also pass@k-particularly on problems where standard RL struggles to make progress. This enables continual improvement over strong open-source models such as DeepScaleR and OpenMath Nemotron, further enhancing their reasoning capabilities. We achieve new state-of-the-art results on math benchmarks using 1.5B-parameter models: 72.50% (+10.73%) on AIME24, 62.29% (+12.79%) on AIME25, and 41.67% (+10.11%) on HMMT25. Code, data and model are available at https://github.com/foreverlasting1202/QuestA.

Targeting sialic acid metabolism: a therapeutic strategy against gastric cancer driven by WZ35

✇Omics in Gastric
By: Boxiang Hong · Ying Wang · Xinran Feng · Lihan Wei · Yichen Huang · Xiawei Zhang · Longyuchen Xu · Pinyan Jin · Liyuan Huang · Rui Li · Xinyu Wu · Zhi Li · Hangyu Ye · Tongke Chen · Xiaoyan Wang · Hao Zhang
23 March 2026 at 18:00

Cell Oncol (Dordr). 2026 Mar 23;49(2):60. doi: 10.1007/s13402-026-01194-6.

ABSTRACT

Glycolytic reprogramming is closely associated with the occurrence and progression of gastric cancer. Specifically, the energy derived from glucose metabolism and the cellular proteins by its intermediate products influence gastric cancer development. However, as an important branch of glucose metabolism, sialic acid metabolism and its mediated sialylation modifications remain insufficiently studied in gastric cancer, and their specific relationship with malignant tumor progression requires further exploration. This study employed a multi‑omics approach, integrating metabolomics, single‑cell RNA sequencing, and bulk RNA sequencing analyses, to investigate the metabolic landscape of gastric cancer and its associated alterations. The results indicated that sialic acid is a characteristic metabolite in malignant gastric cancer tissues. It modulates biological functions such as immune response, proliferative activity, and metabolic remodeling within gastric cancer tissues by influencing sialylation modifications. Furthermore, we identified the drug WZ35, which can inhibit the malignant proliferation of gastric cancer by targeting both sialic acid metabolism and sialylated protein modifications. We put forward a conjecture that the metabolism and modification of sialic acid promote the malignant development of gastric cancer, and we discovered that the drug WZ35 has an inhibitory effect on the sialic acid metabolism of gastric cancer.

GRAPHICAL ABSTRACT:

PMID:41870836 | PMC:PMC13009457 | DOI:10.1007/s13402-026-01194-6

Received — 26 March 2026 ⏭

PERMA: Benchmarking Personalized Memory Agents via Event-Driven Preference and Realistic Task Environments

✇cs.AI, q-bio.NC updates on arXiv.org
By: Shuochen Liu · Junyi Zhu · Long Shu · Junda Lin · Yuhao Chen · Haotian Zhang · Chao Zhang · Derong Xu · Jia Li · Bo Tang · Zhiyu Li · Feiyu Xiong · Enhong Chen · Tong Xu
25 March 2026 at 12:00
arXiv:2603.23231v1 Announce Type: new Abstract: Empowering large language models with long-term memory is crucial for building agents that adapt to users' evolving needs. However, prior evaluations typically interleave preference-related dialogues with irrelevant conversations, reducing the task to needle-in-a-haystack retrieval while ignoring relationships between events that drive the evolution of user preferences. Such settings overlook a fundamental characteristic of real-world personalization: preferences emerge gradually and accumulate across interactions within noisy contexts. To bridge this gap, we introduce PERMA, a benchmark designed to evaluate persona consistency over time beyond static preference recall. Additionally, we incorporate (1) text variability and (2) linguistic alignment to simulate erratic user inputs and individual idiolects in real-world data. PERMA consists of temporally ordered interaction events spanning multiple sessions and domains, with preference-related queries inserted over time. We design both multiple-choice and interactive tasks to probe the model's understanding of persona along the interaction timeline. Experiments demonstrate that by linking related interactions, advanced memory systems can extract more precise preferences and reduce token consumption, outperforming traditional semantic retrieval of raw dialogues. Nevertheless, they still struggle to maintain a coherent persona across temporal depth and cross-domain interference, highlighting the need for more robust personalized memory management in agents. Our code and data are open-sourced at https://github.com/PolarisLiu1/PERMA.

A Synchronous EEG-fNIRS BCI: A Proof-of-Concept for Multimodal Avalanche Analysis of Motor Cognition in Older Adults

✇cs.AI, q-bio.NC updates on arXiv.org
By: Eva Guttmann-Flury · Yun-Hsuan Chen · Qiaoyuan Xiang · Hao Zhang · Mohamad Sawan
25 March 2026 at 12:00
arXiv:2603.23358v1 Announce Type: new Abstract: This proof-of-concept study introduces a novel multimodal framework combining synchronized EEG-fNIRS modalities with neuronal avalanche analysis to identify early network dysfunction in Alzheimer's disease. The approach leverages complementary neural signals to examine motor network dynamics during execution and imagery tasks within an interactive task environment. Preliminary analysis of a small pilot cohort (N=4 subjects, including one with Mild Cognitive Impairment) validated the technical feasibility of the multimodal framework and revealed observable condition-dependent patterns in network organization. Two primary observations emerged: a reduced neural contrast between motor execution and imagery states, and increased trial-to-trial variability in network organization in the MCI participant. These initial results successfully validate the technical pipeline and provide hypothesis-generating observations for future statistically powered studies. The convergence of findings across modalities suggests that multimodal assessment of network flexibility may help detect functional changes in early Alzheimer's continuum, supporting the future development of this BCI-inspired framework into an engaging diagnostic tool.

Mind Your HEARTBEAT! Claw Background Execution Inherently Enables Silent Memory Pollution

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yechao Zhang · Shiqian Zhao · Jie Zhang · Gelei Deng · Jiawen Zhang · Xiaogeng Liu · Chaowei Xiao · Tianwei Zhang
25 March 2026 at 12:00
arXiv:2603.23064v2 Announce Type: cross Abstract: We identify a critical security vulnerability in mainstream Claw personal AI agents: untrusted content encountered during heartbeat-driven background execution can silently pollute agent memory and subsequently influence user-facing behavior without the user's awareness. This vulnerability arises from an architectural design shared across the Claw ecosystem: heartbeat background execution runs in the same session as user-facing conversation, so content ingested from any external source monitored in the background (including email, message channels, news feeds, code repositories, and social platforms) can enter the same memory context used for foreground interaction, often with limited user visibility and without clear source provenance. We formalize this process as an Exposure (E) $\rightarrow$ Memory (M) $\rightarrow$ Behavior (B) pathway: misinformation encountered during heartbeat execution enters the agent's short-term session context, potentially gets written into long-term memory, and later shapes downstream user-facing behavior. We instantiate this pathway in an agent-native social setting using MissClaw, a controlled research replica of Moltbook. We find that (1) social credibility cues, especially perceived consensus, are the dominant driver of short-term behavioral influence, with misleading rates up to 61%; (2) routine memory-saving behavior can promote short-term pollution into durable long-term memory at rates up to 91%, with cross-session behavioral influence reaching 76%; (3) under naturalistic browsing with content dilution and context pruning, pollution still crosses session boundaries. Overall, prompt injection is not required: ordinary social misinformation is sufficient to silently shape agent memory and behavior under heartbeat-driven background execution.
❌