Large language models (LLMs) exhibit substantial potential across diverse scientific disciplines, including materials science. A property prediction framework, ZEBRA‐Prop (Zero‐Shot Embedding‐Based Rapid and Accessible Regression
更新候选
检索、筛选,再回到一手论文。已收录和重复题目自动从此队列排除。
Abstract Long-horizon agents routinely exceed their context windows. A common form of context compaction that reduces a long interaction history to a smaller representation is to summarize the history and continue from the summary
Many languages exhibit differential argument marking (DAM), in which nouns are marked for their grammatical role only in certain contexts. DAM manifests as differential object marking (DOM) and differential subject marking (DSM),
Memory systems enable otherwise-stateless LLM agents to persist user information across sessions, but also introduce a new attack surface. The Trojan Hippo attack is a class of persistent memory attacks that operates under a more
Transformers have evolved from a sequence-to-sequence architecture for machine translation into a general computational framework underlying large language, vision, speech, multimodal, scientific, and decision-making systems. This
During Russia’s full-scale invasion of Ukraine, Telegram became an important platform for both state-aligned communication and opposition-oriented discourse. Over time, anti-Kremlin Telegram channels emerged as prominent venues fo
Abstract Microelectromechanical systems (MEMS) research requires translating device concepts into geometries with physically meaningful behavior under specified structural constraints. Although research agents increasingly automat
Abstract The accelerating penetration of variable renewable energy sources into modern power systems demands digital infrastructures capable of supporting predictive, autonomous, and trustworthy grid operation. While digital twins
Large language models (LLMs) have been extensively used in downstream tasks, and the successful implementation of such tasks typically involves fine-tuning on domain-specific or user-specific data. As these data are generally not
The integration of AI into clinical medicine is an event in the genealogy of medical visibilities where causal-mechanistic and contextual reasoning are required. Extending Dai and Guo (2025) through Pearl’s causal hierarchy, a fou
Cloud-based online services must contend with a load that exhibits significant fluctuations at sub-second granularity. Despite such load volatility, services must meet strict Service Level Objectives (SLOs), which forces a tradeof
The rapid development of AI in journalism, particularly following the emergence of generative AI, has intensified the need for ethical guidance, prompting news organisations and self-regulatory bodies to produce a growing body of
To address the critical challenges of real-time responsiveness and autonomous evolution in highly dynamic flexible production lines, this paper proposes a novel dynamic modeling framework based on a Large Language Model-driven Mul
In time-series forecasting, architectures tailored to domain-specific dataset often outperform generic models. However, manually designing such architectures requires substantial domain knowledge and engineering effort. To address
Self-evolving large language model agents represent a significant shift from static inference systems toward adaptive socio-technical infrastructures that modify their own task repertoires, memory structures, and interaction strat
Large Language Model (LLM) agents are increasingly being used for multi-step reasoning, tool use, information retrieval, and autonomous task execution. However, conventional LLM agents are largely session-oriented: they have limit
The convergence of self-evolution, dynamic memory adaptation, sparsity optimization, and information-geometric principles represents a transformative paradigm in the development of artificial intelligence. Historically, machine le
We introduce and release ScienceBuddy, an interactive scientific research workspace that brings continually improving scientific agents into researchers' everyday workflows. ScienceBuddy supports researchers in carrying out scient
Abstract We study risk-aware self-play in finite-action code repair, separating functional success from test-integrity and forbidden-call violations. From 591 MBPP/HumanEval programs, 395 mutation tasks yield 4,527 candidates. On
Introduction Discharge summary generation is an essential yet time-consuming component of clinical documentation, particularly for post-surgical cancer patients admitted to the Intensive Care Unit (ICU), where clinically relevant
Prompt injection is widely recognized as a major security threat to AI agents that interact with untrusted external data, such as websites, documents, and emails. Prior work has shown that, in the text domain, black-box prompt inj
Agentic AI moves reinforcement learning beyond the single-response setting of RLHF into environments where a model must observe, decide, act, and revise its beliefs over time. This chapter formalizes that shift by recasting LLM-ba
Recursive self-improvement (RSI) seeks to turn gains in an AI system into a stronger capacity to produce subsequent gains. Foundation models make parts of this process practical through generated training data, program revision an
What drove nucleic acids (NA) into functional associations with proteins (PR) at the origins of life? We reason from polymer physics and the Central Dogma that the fitness value of cooperating through a division of labor—NA for re
Abstract LLM-based agents for computational chemistry commonly place the model inside an accumulating execution loop, causing context growth proportional to workflow length, unpredictable token costs, and degraded plan reliability
Real-time embodied social interaction places two equally demanding requirements on an agent: continuously generating fluent multimodal interaction behavior, and proactively reasoning over accumulated dialogue and visual context to
We present Ouroboros, a self-developing agent harness whose tools, prompts, context assembly, and core implementation improve through reviewed commits that become the runtime for later work. Core evolution proceeds in two modes. I
The rapid evolution of Large Language Model (LLM) inference systems has led to significant improvements in serving efficiency. However, our systematic analysis reveals that current evaluation methodologies suffer from fundamental
Deploying million-token Large Language Models (LLMs) is challenging because production workloads are highly heterogeneous, mixing short queries and long documents. This heterogeneity, combined with the quadratic complexity of atte
In threatening situations, animals exhibit a broad range of behavioral and autonomic responses. As such, a crucial adaptive response is the inhibition of pain, which facilitates relevant defensive behaviors that promote survival.
Self-improving coding agents that iteratively rewrite their own source code have demonstrated impressive performance on coding tasks. However, existing solutions generally derive self-modification from a single failure trajectory
Abstract Breast cancer remains a leading cause of cancer morbidity and mortality, representing the most frequently diagnosed new cancer of 2024. Despite overall improvements in survival, women of African ancestry experience dispro
Abstract The rapid integration of artificial intelligence and related emerging technologies has intensified data privacy challenges across the digital economy. Although scholarship has grown substantially, it remains fragmented ac
Abstract Interoception provides a link between bodily states and cognitive and affective processes across both unconscious and conscious domains and is fundamental for a coherent sense of self. While previous research linking inte
The Model Context Protocol (MCP) enables agentic AI frameworks to connect language models to external tool providers, yet its design carries attacker-controlled content, spanning tool descriptions, schema fields, and response data
Abstract Memory has become a central mechanism for extending large language models (LLMs) beyond fixed context windows, stateless interaction, and static parametric knowledge. This survey organizes the literature around three comp
Abstract Migration has been a core topic in German political debate, from postwar expellee displacement to labor migration and recent refugee movements. Large-scale analysis of such political discourse has traditionally required e
Systems with flat band structures, when written in the language of Compact Localised States (CLS), have been shown to be explicitly invariant under supertranslation symmetries, making Carrollian symmetries inherently important for
'Is reason then alone baptized, are the passions pagans?' By placing Edward Young's question on the frontispiece of Either/Or, Søren Kierkegaard announces a problem that runs throughout his authorship: what role do the affects pla
Edge computing has become an important way to support next-generation telecommunications networks. By moving computation closer to data sources and end users, it makes it possible to provide low-latency, energy-efficient, and cont
Abstract Background Tar Spot Complex (TSC) is a devastating foliar disease of maize that threatens global food security, causing yield losses of up to 50%. While the disease is initiated by Phyllachora maydis , the “fisheye” necro
Things are constantly moving; animals, plants, objects, capital, businesses, services, diseases, media, images, information and ideas have mobilities that are dynamic and complex. The Anthropological Handbook of Mobility takes sto
Large Language Model (LLM) agents increasingly serve as personal assistants and workplace collaborators, where their utility depends on memory systems that extract, retrieve, and apply information across long-running conversations
In areas progressing toward malaria elimination, distinguishing locally acquired from imported malaria infections is essential for targeted surveillance. We used whole-genome sequencing (WGS) to characterize Plasmodium vivax infec
Overview & Problem Statement For over half a century, organizational management, public administration, and software engineering have operated under a profound administrative distortion: the linear "Waterfall" model. Originally pr
Abstract Artificial intelligence (AI) agent harnesses have been developed for autonomous scientific simulations, but most current agent harnesses are tested on a single or a small set of software packages. Herein, FermiLink, an ex
Large language models now power autonomous agents capable of complex, multi-step tasks in different environments. Accurate and reliable execution of these tasks requires the agent to predict the results of its actions. Recent rese
The ability of Large Language Models to perform as autonomous conversational agents is fundamentally constrained by the fixed context window of the transformer architecture. Long-term memory mechanisms that approach memory storage
The Catholic Church’s stance toward technology is fundamentally ordered toward integral human development. We are responsible for ensuring that human artifacts and technologies are designed, built, and used in a way that genuinely
The research entitled “Development of Data Structures Using the DataGuard SIEM-DLP Algorithm for Data Loss Prevention and Data Analysis in Distributed Communication Networks” aims to develop the DataGuard SIEM-DLP algorithm. This
AIM: Despite technological advancements, children undergoing adenotonsillectomy using low-temperature plasma radiofrequency ablation remain at risk for postoperative bleeding. This study aimed to construct a precise machine learni
Intelligent resource orchestration in large-scale distributed computing systems can be facilitated via Software-Defined Networking (SDN)-enabled cross-cloud infrastructures, which allow programmability, elasticity, and service-cen
Long-term memory (LTM) is fundamental to large language model (LLM)-based agents in the emerging Internet of Agents (IoA), where distributed multi-agent systems (DMAS) span cloud and edge networks. Existing evaluations are typical
Perceived pay inequity is well established as a predictor of turnover intention, but less is known about how employees respond to it when they stay. Using the exit–voice–loyalty–neglect (EVLN) framework to interpret employees' res
Abstract Much of the current AI safety discourse focuses on deception, value misalignment, and capability discontinuities. We identify a distinct and underappreciated failure mode: the generation of plausible-but-incorrect outputs
Recently, harness self-evolution has emerged as an important paradigm for improving agent performance. Prior studies show that evolving the agent harness can effectively enhance agent capabilities without modifying the foundation
Human decision-making is often influenced by cognitive biases, especially under uncertainty and risk. Prior works have studied biases in single-step decision environments with immediate outcomes or in interaction with a single AI
Abstract Tool-using large language model (LLM) agents exercise authority over protected resources through APIs, tools, credentials, and external services. This structured review asks whether authorization claims are supported at t
The rapid integration of Large Language Models (LLMs) into high-stakes domains necessitates reliable safety and compliance evaluation.However, existing static benchmarks are ill-equipped to address the dynamic nature of AI risks a
Agentic AI moves reinforcement learning beyond the single-response setting of RLHF into environments where a model must observe, decide, act, and revise its beliefs over time. This chapter formalizes that shift by recasting LLM-ba
Large language model (LLM) agents face fundamental limitations in long-horizon reasoning due to finite context windows, making effective memory management critical.Existing methods typically handle long-term memory (LTM) and short
Recursive self-improvement (RSI) asks whether an AI system can improve the process that produces AI systems, so that the next system inherits the improvement. That process is the training algorithm: a better objective or update ru
We present ABACUS, a unified vision-language model that jointly addresses object counting, crowd counting, referring-expression counting, and count-faithful image generation within a single 3B-parameter model. ABACUS introduces th
Abstract Predictive process monitoring enables organizations to anticipate process outcomes from event logs, although deployment in real operational environments remains challenging due to temporal drift, heterogeneous traces, and
The emergence of agentic reinforcement learning (Agentic RL) marks a paradigm shift from conventional reinforcement learning applied to large language models (LLM RL), reframing LLMs from passive sequence generators into autonomou
Interactive digital agents (IDAs) leverage APIs of stateful digital environments to perform tasks in response to user requests. While IDAs powered by instruction-tuned large language models (LLMs) can react to feedback from interf
In dynamic and complex campus environments, robot agents face challenges in providing efficient and accurate services due to their limited ability to learn and adapt to unforeseen situations. To address these issues, we propose Pa
Language model pretraining with next token prediction has proved effective for scaling compute but is limited to the amount of available training data. Scaling reinforcement learning (RL) unlocks a new axis for the continued impro
The rapid advancement of large language models (LLMs) has significantly enhanced the capabilities of agents across various tasks.However, existing agentic systems, whether based on fixed pipeline algorithms or pre-defined meta-lea
Multi-agent systems powered by large language models have demonstrated remarkable capabilities across diverse domains, yet existing automated design approaches seek monolithic solutions that fail to adapt resource allocation based
Background: There is great interest in agentic LLMs, large language models that act as agents. Objectives: We review the growing body of work in this area and provide a research agenda. Methods: Agentic LLMs are LLMs that (1) reas
Large Language Models (LLMs) have demonstrated advanced capabilities in real-world agentic applications. Growing research efforts aim to develop LLM-based agents to address practical demands, introducing a new challenge: agentic s
Large language model (LLM)-based agents have recently attracted much attention from the research and industry communities. Compared with original LLMs, LLM-based agents are featured in their self-evolving capability, which is the
We propose a multi-agent approach (SeM-Agents) based on large language models for medical consultations. This framework incorporates various doctor roles and auxiliary roles, with agents communicating through natural language. Usi
In this paper, we provide a review of the current efforts to develop LLM agents, which are autonomous agents that leverage large language models. We examine the memory management approaches used in these agents. One crucial aspect
In recent years, Large Language Models (LLMs) have shown great abilities in various tasks, including question answering, arithmetic problem solving, and poem writing, among others. Although research on LLM-as-an-agent has shown th
Recent work has embodied LLMs as agents, allowing them to access tools, perform actions, and interact with external content (e.g., emails or websites).However, external content introduces the risk of indirect prompt injection (IPI
Despite the potential of language model-based agents to solve real-world tasks such as web navigation, current methods still struggle with long-horizon tasks with complex action trajectories. In contrast, humans can flexibly solve
Abstract Autonomous agents have long been a research focus in academic and industry communities. Previous research often focuses on training agents with limited knowledge within isolated environments, which diverges significantly