Assisting non-expert users to develop complex interactive websites has become a popular task for LLM-powered code agents. However, existing code agents tend to only generate frontend web pages, masking the lack of real full-stack data processing and storage with fancy visual effects. Notably, constructing production-level full-stack web applications is far more challenging than only generating frontend web pages, demanding careful control of data flow, comprehensive understanding of constantly updating packages and dependencies, and accurate localization of obscure bugs in the codebase. To address these difficulties, we introduce FullStack-Agent, a unified agent system for full-stack agentic coding that consists of three parts: (1) FullStack-Dev, a multi-agent framework with strong planning, code editing, codebase navigation, and bug localization abilities. (2) FullStack-Learn, an innovative data-scaling and self-improving method that back-translates crawled and synthesized website repositories to improve the backbone LLM of FullStack-Dev. (3) FullStack-Bench, a comprehensive benchmark that systematically tests the frontend, backend and database functionalities of the generated website. Our FullStack-Dev outperforms the previous state-of-the-art method by 8.7%, 38.2%, and 15.9% on the frontend, backend, and database test cases respectively. Additionally, FullStack-Learn raises the performance of a 30B model by 9.7%, 9.5%, and 2.8% on the three sets of test cases through self-improvement, demonstrating the effectiveness of our approach. The code is released at https://github.com/mnluzimu/FullStack-Agent.
Long-context inference with Large Language Models (LLMs) is costly due to quadratic attention and growing key-value caches, motivating context compression. In this work, we study soft context compression, where a long context is condensed into a small set of continuous representations. Existing methods typically re-purpose the LLM itself as a trainable compressor, relying on layer-by-layer self-attention to iteratively aggregate information. We argue that this paradigm suffers from two structural limitations: (i) progressive representation overwriting across layers (ii) uncoordinated allocation of compression capacity across tokens. We propose ComprExIT (Context Compression via Explicit Information Transmission), a lightweight framework that formulates soft compression into a new paradigm: explicit information transmission over frozen LLM hidden states. This decouples compression from the model's internal self-attention dynamics. ComprExIT performs (i) depth-wise transmission to selectively transmit multi-layer information into token anchors, mitigating progressive overwriting, and (ii) width-wise transmission to aggregate anchors into a small number of slots via a globally optimized transmission plan, ensuring coordinated allocation of information. Across six question-answering benchmarks, ComprExIT consistently outperforms state-of-the-art context compression methods while introducing only ~1% additional parameters, demonstrating that explicit and coordinated information transmission enables more effective and robust long-context compression.
Recent advances in large language models (LLMs) have made automated multiple-choice question (MCQ) generation increasingly feasible; however, reliably producing items that satisfy controlled cognitive demands remains a challenge. To address this gap, we introduce ReQUESTA, a hybrid, multi-agent framework for generating cognitively diverse MCQs that systematically target text-based, inferential, and main idea comprehension. ReQUESTA decomposes MCQ authoring into specialized subtasks and coordinates LLM-powered agents with rule-based components to support planning, controlled generation, iterative evaluation, and post-processing. We evaluated the framework in a large-scale reading comprehension study using academic expository texts, comparing ReQUESTA-generated MCQs with those produced by a single-pass GPT-5 zero-shot baseline. Psychometric analyses of learner responses assessed item difficulty and discrimination, while expert raters evaluated question quality across multiple dimensions, including topic relevance and distractor quality. Results showed that ReQUESTA-generated items were consistently more challenging, more discriminative, and more strongly aligned with overall reading comprehension performance. Expert evaluations further indicated stronger alignment with central concepts and superior distractor linguistic consistency and semantic plausibility, particularly for inferential questions. These findings demonstrate that hybrid, agentic orchestration can systematically improve the reliability and controllability of LLM-based generation, highlighting workflow design as a key lever for structured artifact generation beyond single-pass prompting.
While existing multi-agent systems (MAS) can handle complex problems by enabling collaboration among multiple agents, they are often highly task-specific, relying on manually crafted agent roles and interaction prompts, which leads to increased architectural complexity and limited reusability across tasks. Moreover, most MAS communicate primarily through natural language, making them vulnerable to error accumulation and instability in long-context, multi-stage interactions within internal agent histories. In this work, we propose \textbf{Agent Primitives}, a set of reusable latent building blocks for LLM-based MAS. Inspired by neural network design, where complex models are built from reusable components, we observe that many existing MAS architectures can be decomposed into a small number of recurring internal computation patterns. Based on this observation, we instantiate three primitives: Review, Voting and Selection, and Planning and Execution. All primitives communicate internally via key-value (KV) cache, which improves both robustness and efficiency by mitigating information degradation across multi-stage interactions. To enable automatic system construction, an Organizer agent selects and composes primitives for each query, guided by a lightweight knowledge pool of previously successful configurations, forming a primitive-based MAS. Experiments show that primitives-based MAS improve average accuracy by 12.0-16.5\% over single-agent baselines, reduce token usage and inference latency by approximately 3$\times$-4$\times$ compared to text-based MAS, while incurring only 1.3$\times$-1.6$\times$ overhead relative to single-agent inference and providing more stable performance across model backbones.
Large Language Models (LLMs) have demonstrated remarkable proficiency in code generation and general reasoning, yet their capacity for autonomous multi-stage planning in high-dimensional, physically constrained environments remains an open research question. This study investigates the limits of current AI agents by evaluating them against the 12th Global Trajectory Optimization Competition (GTOC 12), a complex astrodynamics challenge requiring the design of a large-scale asteroid mining campaign. We adapt the MLE-Bench framework to the domain of orbital mechanics and deploy an AIDE-based agent architecture to autonomously generate and refine mission solutions. To assess performance beyond binary validity, we employ an "LLM-as-a-Judge" methodology, utilizing a rubric developed by domain experts to evaluate strategic viability across five structural categories. A comparative analysis of models, ranging from GPT-4-Turbo to reasoning-enhanced architectures like Gemini 2.5 Pro, and o3, reveals a significant trend: the average strategic viability score has nearly doubled in the last two years (rising from 9.3 to 17.2 out of 26). However, we identify a critical capability gap between strategy and execution. While advanced models demonstrate sophisticated conceptual understanding, correctly framing objective functions and mission architectures, they consistently fail at implementation due to physical unit inconsistencies, boundary condition errors, and inefficient debugging loops. We conclude that, while current LLMs often demonstrate sufficient knowledge and intelligence to tackle space science tasks, they remain limited by an implementation barrier, functioning as powerful domain facilitators rather than fully autonomous engineers.
Recent progress has expanded the use of large language models (LLMs) in drug discovery, including synthesis planning. However, objective evaluation of retrosynthesis performance remains limited. Existing benchmarks and metrics typically rely on published synthetic procedures and Top-K accuracy based on single ground-truth, which does not capture the open-ended nature of real-world synthesis planning. We propose a new benchmarking framework for single-step retrosynthesis that evaluates both general-purpose and chemistry-specialized LLMs using ChemCensor, a novel metric for chemical plausibility. By emphasizing plausibility over exact match, this approach better aligns with human synthesis planning practices. We also introduce CREED, a novel dataset comprising millions of ChemCensor-validated reaction records for LLM training, and use it to train a model that improves over the LLM baselines under this benchmark.
Large language models (LLMs) are trained and tested extensively on symbolic representations such as code and graphs, yet real-world user tasks are often specified in natural language. To what extent can LLMs generalize across these representations? Here, we approach this question by studying isomorphic tasks involving procedures represented in code, graphs, and natural language (e.g., scheduling steps in planning). We find that training LLMs with popular post-training methods on graphs or code data alone does not reliably generalize to corresponding natural language tasks, while training solely on natural language can lead to inefficient performance gains. To address this gap, we propose a two-stage data curriculum that first trains on symbolic, then natural language data. The curriculum substantially improves model performance across model families and tasks. Remarkably, a 1.5B Qwen model trained by our method can closely match zero-shot GPT-4o in naturalistic planning. Finally, our analysis suggests that successful cross-representation generalization can be interpreted as a form of generative analogy, which our curriculum effectively encourages.
Open-domain multimodal document retrieval aims to retrieve specific components (paragraphs, tables, or images) from large and interconnected document corpora. Existing graph-based retrieval approaches typically rely on a uniform similarity metric that overlooks hop-specific semantics, and their rigid pre-defined plans hinder dynamic error correction. These limitations suggest that a retriever should adapt its reasoning to the evolving context and recover intelligently from dead ends. To address these needs, we propose Failure is Feedback (FiF), which casts subgraph retrieval as a sequential decision process and introduces two key innovations. (i) We introduce a history-aware backtracking mechanism; unlike standard backtracking that simply reverts the state, our approach piggybacks on the context of failed traversals, leveraging insights from previous failures. (ii) We implement an economically-rational agentic workflow. Unlike conventional agents with static strategies, our orchestrator employs a cost-aware traversal method to dynamically manage the trade-off between retrieval accuracy and inference costs, escalating to intensive LLM-based reasoning only when the prior failure justifies the additional computational investment. Extensive experiments show that FiF achieves state-of-the-art retrieval on the benchmarks of MultimodalQA, MMCoQA and WebQA.
Computer-use agents (CUAs) that interact with real computer systems can perform automated tasks but face critical safety risks. Ambiguous instructions may trigger harmful actions, and adversarial users can manipulate tool execution to achieve malicious goals. Existing benchmarks mostly focus on short-horizon or GUI-based tasks, evaluating on execution-time errors but overlooking the ability to anticipate planning-time risks. To fill this gap, we present LPS-Bench, a benchmark that evaluates the planning-time safety awareness of MCP-based CUAs under long-horizon tasks, covering both benign and adversarial interactions across 65 scenarios of 7 task domains and 9 risk types. We introduce a multi-agent automated pipeline for scalable data generation and adopt an LLM-as-a-judge evaluation protocol to assess safety awareness through the planning trajectory. Experiments reveal substantial deficiencies in existing CUAs' ability to maintain safe behavior. We further analyze the risks and propose mitigation strategies to improve long-horizon planning safety in MCP-based CUA systems. We open-source our code at https://github.com/tychenn/LPS-Bench.
As code large language models (LLMs) evolve into tool-interactive agents via the Model Context Protocol (MCP), their generalization is increasingly limited by low-quality synthetic data and the diminishing returns of quantity scaling. Moreover, quantity-centric scaling exhibits an early bottleneck that underutilizes trajectory data. We propose TDScaling, a Trajectory Diversity Scaling-based data synthesis framework for code agents that scales performance through diversity rather than raw volume. Under a fixed training budget, increasing trajectory diversity yields larger gains than adding more trajectories, improving the performance-cost trade-off for agent training. TDScaling integrates four innovations: (1) a Business Cluster mechanism that captures real-service logical dependencies; (2) a blueprint-driven multi-agent paradigm that enforces trajectory coherence; (3) an adaptive evolution mechanism that steers synthesis toward long-tail scenarios using Domain Entropy, Reasoning Mode Entropy, and Cumulative Action Complexity to prevent mode collapse; and (4) a sandboxed code tool that mitigates catastrophic forgetting of intrinsic coding capabilities. Experiments on general tool-use benchmarks (BFCL, tau^2-Bench) and code agent tasks (RebenchT, CodeCI, BIRD) demonstrate a win-win outcome: TDScaling improves both tool-use generalization and inherent coding proficiency. We plan to release the full codebase and the synthesized dataset (including 30,000+ tool clusters) upon publication.
AI-powered planning tools show promise in supporting programming learners by enabling early, formative feedback on their thinking processes prior to coding. To date, however, most AI-supported planning tools rely on students' natural-language explanations, using LLMs to interpret learners' descriptions of their algorithmic intent. Prior to the emergence of LLM-based systems, CS education research extensively studied trace-based planning in pen-and-paper settings, demonstrating that reasoning through stepwise execution with explicit state transitions helps learners build and refine mental models of program behavior. Despite its potential, little is known about how tracing interacts with AI-mediated feedback and whether integrating tracing into AI-supported planning tools leads to different learning processes or interaction dynamics compared to natural-language-based planning alone. We study how requiring learners to produce explicit execution traces with an AI-supported planning tool affects their algorithmic reasoning. In a between-subjects study with 20 students, tracing shifted learners away from code-like, line-by-line descriptions toward more goal-driven reasoning about program behavior. Moreover, it led to more consistent partially correct solutions, although final coding performance remained comparable across conditions. Notably, tracing did not significantly affect the quality or reliability of LLM-generated feedback. These findings reveal tradeoffs in combining tracing with AI-supported planning and inform design guidelines for integrating natural language, tracing, and coding to support iterative reasoning throughout the programming process.
Large language models (LLMs) have enabled a new class of agentic AI systems that reason, plan, and act by invoking external tools. However, most existing agentic architectures remain centralized and monolithic, limiting scalability, specialization, and interoperability. This paper proposes a framework for scalable agentic intelligence, termed the Internet of Agentic AI, in which autonomous, heterogeneous agents distributed across cloud and edge infrastructure dynamically form coalitions to execute task-driven workflows. We formalize a network-native model of agentic collaboration and introduce an incentive-compatible workflow-coalition feasibility framework that integrates capability coverage, network locality, and economic implementability. To enable scalable coordination, we formulate a minimum-effort coalition selection problem and propose a decentralized coalition formation algorithm. The proposed framework can operate as a coordination layer above the Model Context Protocol (MCP). A healthcare case study demonstrates how domain specialization, cloud-edge heterogeneity, and dynamic coalition formation enable scalable, resilient, and economically viable agentic workflows. This work lays the foundation for principled coordination and scalability in the emerging era of Internet of Agentic AI.
Group Relative Policy Optimization (GRPO) has recently emerged as a practical recipe for aligning large language models with verifiable objectives. However, under sparse terminal rewards, GRPO often stalls because rollouts within a group frequently receive identical rewards, causing relative advantages to collapse and updates to vanish. We propose self-hint aligned GRPO with privileged supervision (SAGE), an on-policy reinforcement learning framework that injects privileged hints during training to reshape the rollout distribution under the same terminal verifier reward. For each prompt $x$, the model samples a compact hint $h$ (e.g., a plan or decomposition) and then generates a solution $τ$ conditioned on $(x,h)$. Crucially, the task reward $R(x,τ)$ is unchanged; hints only increase within-group outcome diversity under finite sampling, preventing GRPO advantages from collapsing under sparse rewards. At test time, we set $h=\varnothing$ and deploy the no-hint policy without any privileged information. Moreover, sampling diverse self-hints serves as an adaptive curriculum that tracks the learner's bottlenecks more effectively than fixed hints from an initial policy or a stronger external model. Experiments over 6 benchmarks with 3 LLMs show that SAGE consistently outperforms GRPO, on average +2.0 on Llama-3.2-3B-Instruct, +1.2 on Qwen2.5-7B-Instruct and +1.3 on Qwen3-4B-Instruct. The code is available at https://github.com/BaohaoLiao/SAGE.
Multi-agent LLM frameworks are widely used to accelerate the development of agent systems powered by large language models (LLMs). These frameworks impose distinct architectural structures that govern how agents interact, store information, and coordinate tasks. However, their impact on system performance remains poorly understood. This gap is critical, as architectural choices alone can induce order-of-magnitude differences in latency and throughput, as well as substantial variation in accuracy and scalability. Addressing this challenge requires (i) jointly evaluating multiple capabilities, such as orchestration overhead, memory behavior, planning, specialization, and coordination, and (ii) conducting these evaluations under controlled, framework-level conditions to isolate architectural effects. Existing benchmarks focus on individual capabilities and lack standardized framework-level evaluation. We address these limitations by (i) introducing an architectural taxonomy for systematically comparing multi-agent LLM frameworks along fundamental dimensions, and (ii) developing MAFBench, a unified evaluation suite that integrates existing benchmarks under a standardized execution pipeline. Using MAFBench, we conduct a controlled empirical study across several widely used frameworks. Our results show that framework-level design choices alone can increase latency by over 100x, reduce planning accuracy by up to 30%, and lower coordination success from above 90% to below 30%. Finally, we translate our findings into concrete architectural design principles and framework selection guidance, and outline promising future research directions.
Large language models (LLMs) have been shown to acquire sequence-level planning abilities during training, yet their planning behavior exhibited at inference time often appears short-sighted and inconsistent with these capabilities. We propose a Bayesian account for this gap by grounding planning behavior in the evolving generative context: given the subtle differences between natural language and the language internalized by LLMs, accumulated self-generated context drives a planning-shift during inference and thereby creates the appearance of compromised planning behavior. We further validate the proposed model through two controlled experiments: a random-generation task demonstrating constrained planning under human prompts and increasing planning strength as self-generated context accumulates, and a Gaussian-sampling task showing reduced initial bias when conditioning on self-generated sequences. These findings provide a theoretical explanation along with empirical evidence for characterizing how LLMs plan ahead during inference.
Autonomous agents powered by large language models (LLMs) promise to accelerate scientific discovery end-to-end, but rigorously evaluating their capacity for verifiable discovery remains a central challenge. Existing benchmarks face a trade-off: they either heavily rely on LLM-as-judge evaluations of automatically generated research outputs or optimize convenient yet isolated performance metrics that provide coarse proxies for scientific insight. To address this gap, we introduce FIRE-Bench (Full-cycle Insight Rediscovery Evaluation), a benchmark that evaluates agents through the rediscovery of established findings from recent, high-impact machine learning research. Agents are given only a high-level research question extracted from a published, verified study and must autonomously explore ideas, design experiments, implement code, execute their plans, and derive conclusions supported by empirical evidence. We evaluate a range of state-of-the-art agents with frontier LLMs backbones like gpt-5 on FIRE-Bench. Our results show that full-cycle scientific research remains challenging for current agent systems: even the strongest agents achieve limited rediscovery success (<50 F1), exhibit high variance across runs, and display recurring failure modes in experimental design, execution, and evidence-based reasoning. FIRE-Bench provides a rigorous and diagnostic framework for measuring progress toward reliable agent-driven scientific discovery.
Telecommunications networks generate extensive performance and environmental telemetry, yet most LTE and 5G-NR deployments still rely on static, manually engineered configurations. This limits adaptability in rural, nomadic, and bandwidth-constrained environments where traffic distributions, propagation characteristics, and user behavior fluctuate rapidly. Artificial Intelligence (AI), more specifically Machine Learning (ML) models, provide new opportunities to transition Radio Access Networks (RANs) from rigid, rule-based systems toward adaptive, self-optimizing infrastructures that can respond autonomously to these dynamics. This paper proposes a practical architecture incorporating AI-assisted planning, reinforcement-learning-based RAN optimization, real-time telemetry analytics, and digital-twin-based validation. In parallel, the paper addresses the challenge of delivering embodied-AI healthcare services, educational tools, and large language model (LLM) applications to communities with insufficient backhaul for cloud computing. We introduce an edge-hosted execution model in which applications run directly on LTE/5G-NR base stations using containers, reducing latency and bandwidth consumption while improving resilience. Together, these contributions demonstrate how AI can enhance network performance, reduce operational overhead, and expand access to advanced digital services, aligning with broader goals of sustainable and inclusive network development.
Large language models are increasingly deployed as specialized agents that plan, call tools, and take actions over extended horizons. Yet many existing evaluations assume a "clean interface" where dynamics are specified and stable, tools and sensors are reliable, and success is captured by a single explicit objective-often overestimating real-world readiness. In practice, agents face underspecified rules, unreliable signals, shifting environments, and implicit, multi-stakeholder goals. The challenge is therefore not just solving tasks, but adapting while solving: deciding what to trust, what is wanted, when to verify, and when to fall back or escalate. We stress-test deployment-relevant robustness under four operational circumstances: partial observability, dynamic environments, noisy signals, and dynamic agent state. We benchmark agentic LLMs in a grid-based game with a simple goal but long-horizon execution. Episodes violate clean-interface assumptions yet remain solvable, forcing agents to infer rules, pay for information, adapt to environmental and internal shifts, and act cautiously under noise. Across five state-of-the-art LLM agents, we find large gaps between nominal task-solving and deployment-like robustness. Performance generally degrades as grid size and horizon increase, but rankings are unstable: weaker models can beat stronger ones when strategy matches the uncertainty regime. Despite no explicit instruction, agents trade off completion, efficiency, and penalty avoidance, suggesting partial objective inference. Ablations and feature analyses reveal model-specific sensitivities and failure drivers, motivating work on verification, safe action selection, and objective inference under partial observability, noise, and non-stationarity.
Automating AI research differs from general software engineering due to computationally expensive evaluation (e.g., model training) and opaque performance attribution. Current LLM-based agents struggle here, often generating monolithic scripts that ignore execution costs and causal factors. We introduce MARS (Modular Agent with Reflective Search), a framework optimized for autonomous AI research. MARS relies on three pillars: (1) Budget-Aware Planning via cost-constrained Monte Carlo Tree Search (MCTS) to explicitly balance performance with execution expense; (2) Modular Construction, employing a "Design-Decompose-Implement" pipeline to manage complex research repositories; and (3) Comparative Reflective Memory, which addresses credit assignment by analyzing solution differences to distill high-signal insights. MARS achieves state-of-the-art performance among open-source frameworks on MLE-Bench under comparable settings, maintaining competitiveness with the global leaderboard's top methods. Furthermore, the system exhibits qualitative "Aha!" moments, where 63% of all utilized lessons originate from cross-branch transfer, demonstrating that the agent effectively generalizes insights across search paths.
LLM-based deep research agents are largely built on the ReAct framework. This linear design makes it difficult to revisit earlier states, branch into alternative search directions, or maintain global awareness under long contexts, often leading to local optima, redundant exploration, and inefficient search. We propose Re-TRAC, an agentic framework that performs cross-trajectory exploration by generating a structured state representation after each trajectory to summarize evidence, uncertainties, failures, and future plans, and conditioning subsequent trajectories on this state representation. This enables iterative reflection and globally informed planning, reframing research as a progressive process. Empirical results show that Re-TRAC consistently outperforms ReAct by 15-20% on BrowseComp with frontier LLMs. For smaller models, we introduce Re-TRAC-aware supervised fine-tuning, achieving state-of-the-art performance at comparable scales. Notably, Re-TRAC shows a monotonic reduction in tool calls and token usage across rounds, indicating progressively targeted exploration driven by cross-trajectory reflection rather than redundant search.
Large language models can generate fluent explanations, but effective tutoring requires supporting the learner's thought process, not just delivering content. Metacognitive tutoring targets this gap by prompting planning, monitoring, debugging, and evaluation, and crucially, deciding when to be active versus minimally present, based on learner signals and trajectory. We introduce MetaCLASS, a learning-science grounded framework that formulates metacognitive tutoring as move selection over 11 interpretable actions aligned to self-regulated learning processes. MetaCLASS uses a two-phase framework that first plans a pedagogical trajectory conditioned on learner profiles (calibration, help-seeking) and then generates natural dialogue consistent with that plan. This yields a dataset of 1,015 conversations (7,711 turns) annotated with turn-level metacognitive labels, and validated for pedagogical contingency and trajectory adherence. We benchmark nine LLMs on predicting the next coach move given the problem and dialogue context. The best model achieves only 43.2\% accuracy, and models exhibit compulsive intervention bias: in turns where effective metacognitive tutoring requires silent (41.7\% of cases), models predict `no intervention' only 4.2\% of the time, while severely over-predicting high-intervention moves. These results show that traditional content-based tutoring ability does not translate to metacognitive tutoring competence, positioning MetaCLASS as a testbed for developing intelligent tutors that promote self-regulated learning.
Large language models (LLMs) require enormous computing power to pretrain on massive datasets. When limited datasets are available, smaller-sized LLMs are better choice to pretrain (on user-specified datasets) by following the scaling laws of LLMs. Using pretrained models, vector embeddings can be generated for raw data and stored using vector databases to support modern AI applications and semantic search. In this work, we investigate the performance of pretraining techniques for smaller-sized LLMs on an experimental testbed (with commodity GPUs) available to academic users at no charge. We consider data parallelism, intra-operator parallelism, and inter-operator/pipeline parallelism, and their combinations for pretraining. We set up different GPU clusters with homogeneous and heterogeneous GPU hardware. Furthermore, we investigate the impact of network latency on pretraining performance especially when GPUs are geographically distributed. We used GPT-2 medium and large models and pretrained them using open-source packages, namely, Alpa and Ray. We observed that Alpa's execution plans that collectively optimized intra-operator and inter-operator/pipeline parallelism consistently performed the best when GPUs were geographically distributed. This was especially true when the network latencies were in 10's of milliseconds. Based on the insights gained from the experiments, we propose a systematic approach for selecting the appropriate pretraining technique to achieve high training performance/lower execution time as well as to reduce the number of GPUs used.
Large language models (LLMs) have shown promise in assisting cybersecurity tasks, yet existing approaches struggle with automatic vulnerability discovery and exploitation due to limited interaction, weak execution grounding, and a lack of experience reuse. We propose Co-RedTeam, a security-aware multi-agent framework designed to mirror real-world red-teaming workflows by integrating security-domain knowledge, code-aware analysis, execution-grounded iterative reasoning, and long-term memory. Co-RedTeam decomposes vulnerability analysis into coordinated discovery and exploitation stages, enabling agents to plan, execute, validate, and refine actions based on real execution feedback while learning from prior trajectories. Extensive evaluations on challenging security benchmarks demonstrate that Co-RedTeam consistently outperforms strong baselines across diverse backbone models, achieving over 60% success rate in vulnerability exploitation and over 10% absolute improvement in vulnerability detection. Ablation and iteration studies further confirm the critical role of execution feedback, structured interaction, and memory for building robust and generalizable cybersecurity agents.
This work stems from prior complementary observations on the dynamics of Chain-of-Thought (CoT): Large Language Models (LLMs) is shown latent planning of subsequent reasoning prior to CoT emergence, thereby diminishing the significance of explicit CoT; whereas CoT remains critical for tasks requiring multi-step reasoning. To deepen the understanding between LLM's internal states and its verbalized reasoning trajectories, we investigate the latent planning strength of LLMs, through our probing method, Tele-Lens, applying to hidden states across diverse task domains. Our empirical results indicate that LLMs exhibit a myopic horizon, primarily conducting incremental transitions without precise global planning. Leveraging this characteristic, we propose a hypothesis on enhancing uncertainty estimation of CoT, which we validate that a small subset of CoT positions can effectively represent the uncertainty of the entire path. We further underscore the significance of exploiting CoT dynamics, and demonstrate that automatic recognition of CoT bypass can be achieved without performance degradation. Our code, data and models are released at https://github.com/lxucs/tele-lens.
Large language models (LLMs) increasingly support heterogeneous tasks within a single interface, requiring users to form, update, and act upon beliefs about one system across domains with different reliability profiles. Understanding how such beliefs transfer across tasks and shape delegation is therefore critical for the design of multipurpose AI systems. We report a preregistered experiment (N=240; 7,200 trials) in which participants interacted with a controlled AI simulation across grammar checking, travel planning, and visual question answering, each with fixed, domain-typical accuracy levels. Delegation was operationalized as a binary reliance decision: accepting the AI's output versus acting independently, and belief dynamics were evaluated against Bayesian benchmarks. We find three main results. First, participants do not reset beliefs between tasks: priors in a new task depend on posteriors from the previous task, with a 10-point increase predicting a 3-4 point higher subsequent prior. Second, within tasks, belief updating follows the Bayesian direction but is substantially conservative, proceeding at roughly half the normative Bayesian rate. Third, delegation is driven primarily by subjective beliefs about AI accuracy rather than self-confidence, though confidence independently reduces reliance when beliefs are held constant. Together, these findings show that users form global, path-dependent expectations about multipurpose AI systems, update them conservatively, and rely on AI primarily based on subjective beliefs rather than objective performance. We discuss implications for expectation calibration, reliance design, and the risks of belief spillovers in deployed LLM-based interfaces.
Recent advances in Large Language Models (LLMs) and Vision-Language Models (VLMs) have enabled powerful semantic and multimodal reasoning capabilities, creating new opportunities to enhance sample efficiency, high-level planning, and interpretability in reinforcement learning (RL). While prior work has integrated LLMs and VLMs into various components of RL, the replay buffer, a core component for storing and reusing experiences, remains unexplored. We propose addressing this gap by leveraging VLMs to guide the prioritization of experiences in the replay buffer. Our key idea is to use a frozen, pre-trained VLM (requiring no fine-tuning) as an automated evaluator to identify and prioritize promising sub-trajectories from the agent's experiences. Across scenarios, including game-playing and robotics, spanning both discrete and continuous domains, agents trained with our proposed prioritization method achieve 11-52% higher average success rates and improve sample efficiency by 19-45% compared to previous approaches. https://esharony.me/projects/vlm-rb/
Recent advances in robot learning increasingly rely on LLM-based task planning, leveraging their ability to bridge natural language with executable actions. While prior works showcased great performances, the widespread adoption of these models in robotics has been challenging as 1) existing methods are often closed-source or computationally intensive, neglecting the actual deployment on real-world physical systems, and 2) there is no universally accepted, plug-and-play representation for robotic task generation. Addressing these challenges, we propose BTGenBot-2, a 1B-parameter open-source small language model that directly converts natural language task descriptions and a list of robot action primitives into executable behavior trees in XML. Unlike prior approaches, BTGenBot-2 enables zero-shot BT generation, error recovery at inference and runtime, while remaining lightweight enough for resource-constrained robots. We further introduce the first standardized benchmark for LLM-based BT generation, covering 52 navigation and manipulation tasks in NVIDIA Isaac Sim. Extensive evaluations demonstrate that BTGenBot-2 consistently outperforms GPT-5, Claude Opus 4.1, and larger open-source models across both functional and non-functional metrics, achieving average success rates of 90.38% in zero-shot and 98.07% in one-shot, while delivering up to 16x faster inference compared to the previous BTGenBot.
As LLM-based agents are deployed in increasingly complex real-world settings, existing benchmarks underrepresent key challenges such as enforcing global constraints, coordinating multi-tool reasoning, and adapting to evolving user behavior over long, multi-turn interactions. To bridge this gap, we introduce \textbf{TRIP-Bench}, a long-horizon benchmark grounded in realistic travel-planning scenarios. TRIP-Bench leverages real-world data, offers 18 curated tools and 40+ travel requirements, and supports automated evaluation. It includes splits of varying difficulty; the hard split emphasizes long and ambiguous interactions, style shifts, feasibility changes, and iterative version revision. Dialogues span up to 15 user turns, can involve 150+ tool calls, and may exceed 200k tokens of context. Experiments show that even advanced models achieve at most 50\% success on the easy split, with performance dropping below 10\% on hard subsets. We further propose \textbf{GTPO}, an online multi-turn reinforcement learning method with specialized reward normalization and reward differencing. Applied to Qwen2.5-32B-Instruct, GTPO improves constraint satisfaction and interaction robustness, outperforming Gemini-3-Pro in our evaluation. We expect TRIP-Bench to advance practical long-horizon interactive agents, and GTPO to provide an effective online RL recipe for robust long-horizon training.
While large language models have become the prevailing approach for agentic reasoning and planning, their success in symbolic domains does not readily translate to the physical world. Spatial intelligence, the ability to perceive 3D structure, reason about object relationships, and act under physical constraints, is an orthogonal capability that proves important for embodied agents. Existing surveys address either agentic architectures or spatial domains in isolation. None provide a unified framework connecting these complementary capabilities. This paper bridges that gap. Through a thorough review of over 2,000 papers, citing 742 works from top-tier venues, we introduce a unified three-axis taxonomy connecting agentic capabilities with spatial tasks across scales. Crucially, we distinguish spatial grounding (metric understanding of geometry and physics) from symbolic grounding (associating images with text), arguing that perception alone does not confer agency. Our analysis reveals three key findings mapped to these axes: (1) hierarchical memory systems (Capability axis) are important for long-horizon spatial tasks. (2) GNN-LLM integration (Task axis) is a promising approach for structured spatial reasoning. (3) World models (Scale axis) are essential for safe deployment across micro-to-macro spatial scales. We conclude by identifying six grand challenges and outlining directions for future research, including the need for unified evaluation frameworks to standardize cross-domain assessment. This taxonomy provides a foundation for unifying fragmented research efforts and enabling the next generation of spatially-aware autonomous systems in robotics, autonomous vehicles, and geospatial intelligence.
Modern scientific research relies on large-scale data, complex workflows, and specialized tools, which existing LLMs and tool-based agents struggle to handle due to limitations in long-horizon planning, robust goal maintenance, and continual learning from execution. To address these issues, in this work, we propose S1-NexusAgent, a self-evolving agent framework designed for multidisciplinary scientific research. S1-NexusAgent adopts a hierarchical Plan-and-CodeAct execution paradigm, decoupling global scientific planning from subtask-level tool execution through a dual-loop architecture, thereby enabling stable modeling of complex research workflows. The system natively supports the Model Context Protocol (MCP), integrates up to thousands of cross-disciplinary scientific tools, and achieves efficient orchestration of heterogeneous research tools via intention-aware dynamic tool retrieval and hot-plug mechanisms. To address long-context and large-scale data challenges in scientific settings, S1-NexusAgent introduces object-reference-based sparse context management, which enables sub-task context isolation and intermediate result compression. Building on this, a Critic Agent automatically evaluates complete execution trajectories and distills high-quality research paths into reusable Scientific Skills, forming a closed loop for continuous self-evolution, which is valuable for sustainable and long-horizon scientific research. Experiments on authoritative scientific benchmarks involving long-horizon planning and complex specialized tool orchestration, including biomini-eval (biology), ChemBench (chemistry), and MatSciBench (material science), demonstrate that S1-NexusAgent achieves state-of-the-art performance, validating its effectiveness and generalization capability in complex scientific tasks.