Search: “tools”

arXiv:2512.22673v3 Announce Type: replace Abstract: Travel planning is a natural real-world task to test large language models' (LLMs) planning and tool-use abilities. Although prior work has studied

arXiv:2604.17256v1 Announce Type: cross Abstract: Assessing the security posture of modern computing systems typically requires the use of multiple specialized tools. These tools focus on different as

arXiv:2502.11271v2 Announce Type: replace-cross Abstract: Solving complex reasoning tasks may involve visual understanding, domain knowledge retrieval, numerical calculation, and multi-step reasoning.

arXiv:2604.10261v1 Announce Type: new Abstract: Existing tool-use benchmarks for LLM agents are overwhelmingly linear: our analysis of six benchmarks shows 55 to 100% of instances are simple chains of

arXiv:2604.11557v1 Announce Type: new Abstract: Tool-use capability is a fundamental component of LLM agents, enabling them to interact with external systems through structured function calls. However

arXiv:2607.27703v2 Announce Type: replace Abstract: Vision-language models (VLMs) are increasingly used in embodied agents to interpret visual inputs, reason about spatial relationships, and make task

arXiv:2608.04007v1 Announce Type: cross Abstract: Tool-Integrated Reasoning (TIR) enables LLMs to solve complex tasks through iterative tool interactions. However, existing reinforcement learning meth

arXiv:2607.27083v1 Announce Type: new Abstract: As LLM agents increasingly depend on diverse external services such as search engines, databases, and connectors, agent harnesses face a fundamental too

arXiv:2607.17528v3 Announce Type: replace Abstract: Large language model (LLM) agents are extending electronic design automation (EDA) beyond static RTL generation toward long-horizon, tool-interactiv

arXiv:2607.17075v2 Announce Type: replace-cross Abstract: The advent of LLMs has significantly changed the research on privacy policy and data compliance analysis by enabling tasks that previously req

arXiv:2607.05465v1 Announce Type: cross Abstract: Complex image creation and editing often require more than a single generation or editing model. A user request may involve synthesizing images, local

arXiv:2607.05780v1 Announce Type: cross Abstract: While humans readily repurpose a book, a stone, or a shoe to drive a nail, robots trained on specific tools fail to transfer the same function to nove

arXiv:2512.07287v3 Announce Type: replace-cross Abstract: As intents unfold and environments change, multi-turn agents face continuously shifting decision contexts. Although reusing past experience is

arXiv:2601.05366v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly deployed as agents that invoke external tools through structured function calls. While recent wo

arXiv:2606.26474v1 Announce Type: cross Abstract: Fine-tuning through RL reshapes the internal representations of language models to enable agentic behaviors such as tool use, yet the mechanistic basi

arXiv:2606.26027v1 Announce Type: new Abstract: Tool use enables large language models (LLMs) to perform complex tasks, and recent agentic reinforcement learning (RL) methods show promise for enhancin

arXiv:2606.23112v1 Announce Type: new Abstract: Multi-turn tool-using agents must coordinate long-horizon tool sequences while tracking dialogue state and policy constraints. Existing approaches often

arXiv:2602.04935v3 Announce Type: replace-cross Abstract: Adapting LLM agents to domain-specific tool calling remains notably brittle under evolving interfaces. Prompt and schema engineering is easy t

arXiv:2606.06976v1 Announce Type: new Abstract: Large language model (LLM)-based agents often make suboptimal tool-use decisions, including unsupported tool invocation and hallucinated direct response

arXiv:2606.05784v1 Announce Type: new Abstract: We identify and formally characterize credit misassignment as a systematic failure mode of GRPO in tool-augmented multimodal search agents: its uniform

arXiv:2606.05806v1 Announce Type: new Abstract: Existing benchmarks evaluate Tool-Integrated Reasoning (TIR) in LLMs on idealized ''happy paths'', largely overlooking real-world tool failures. We intr

arXiv:2606.05728v1 Announce Type: cross Abstract: Generating executable tool plans requires selecting appropriate subsets from tool libraries, a combinatorial search problem with an exponentially larg

arXiv:2602.14849v2 Announce Type: replace-cross Abstract: LLM agents execute multi-step workflows that mutate external state through tools. Common orchestrators treat tool return as the settlement tri

arXiv:2606.00135v1 Announce Type: cross Abstract: Tool-calling is a central component of modern large language model (LLM) agents, equipping them with skills beyond their parametric knowledge. This pa

arXiv:2605.29568v1 Announce Type: new Abstract: Tool-Integrated Reasoning (TIR) extends LLM capabilities by leveraging external environments. However, existing methods lack the deliberation during seq

arXiv:2605.26037v1 Announce Type: new Abstract: We test the standard RLVR tool-use recipe -- GRPO on Qwen2.5-7B-Instruct -- on a deliberately minimal knowledge-graph tool API: four Freebase navigation

arXiv:2507.10593v3 Announce Type: replace-cross Abstract: Every LLM tool call is structurally an RPC -- a function name, JSON arguments, and a serialized result -- yet each protocol (native Python, MC

arXiv:2605.13245v1 Announce Type: new Abstract: Language models can produce convincing scientific analyses, but repeated generations on the same data do not guarantee the same result. A researcher may

arXiv:2605.09679v1 Announce Type: cross Abstract: Medical vision-language models (VLMs) and AI agents have made significant progress in learning to analyze and reason about clinical images. However, e

arXiv:2601.18700v2 Announce Type: replace Abstract: Emotional Support Conversation requires not only affective expression but also grounded instrumental support to provide trustworthy guidance. Howeve

arXiv:2605.07339v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities in orchestrating tools for reasoning tasks. However, existing methods rely on a s

arXiv:2605.02964v1 Announce Type: new Abstract: Reinforcement learning (RL) trained language model agents with tool access are increasingly deployed in coding assistants, research tools, and autonomou

arXiv:2605.00876v1 Announce Type: cross Abstract: Vision-language models (VLMs) read an image and produce text in a single forward pass, whereas radiologists typically inspect an image several times a

arXiv:2604.27233v1 Announce Type: new Abstract: Tool-calling agents are evaluated on tool selection, parameter accuracy, and scope recognition, yet LLM trajectory assessments remain inherently post-ho

arXiv:2604.17739v1 Announce Type: cross Abstract: Reinforcement learning (RL) has become a prevalent paradigm for training tool calling agents, which typically requires online interactive environments

arXiv:2508.08791v3 Announce Type: replace Abstract: Effective tool use is essential for large language models (LLMs) to interact with their environment. However, progress is limited by the lack of eff

arXiv:2604.08545v1 Announce Type: new Abstract: The advent of agentic multimodal models has empowered systems to actively interact with external environments. However, current agents suffer from a pro

arXiv:2608.10039v1 Announce Type: new Abstract: Agentic workflows have become an important abstraction for building reliable LLM-based automation systems by organizing large language models (LLMs), to

arXiv:2608.10827v1 Announce Type: cross Abstract: Medical visual agents can use tools to inspect images and retrieve external knowledge, but indiscriminate tool use may introduce noisy or misleading e

arXiv:2608.07959v1 Announce Type: new Abstract: Ultra-long egocentric video understanding requires reasoning over temporally sparse evidence distributed across hours or days, challenging current multi

arXiv:2608.07955v1 Announce Type: new Abstract: Large vision-language models have achieved strong performance in multimodal reasoning, but they remain unreliable on fine-grained spatial tasks that dem

arXiv:2608.08048v1 Announce Type: cross Abstract: This paper presents, for the first time in power systems literature to our knowledge, analytical tools to explain the training performance of machine

arXiv:2608.02110v1 Announce Type: new Abstract: Executing long-horizon tool invocations in real-world environments is severely challenged by dynamic user intent noise. Existing methods attempt robustn

arXiv:2608.00814v1 Announce Type: new Abstract: LLMs generate tool calls token by token, even though the function choice and argument values can often be predicted in parallel from the request and too

arXiv:2607.28595v1 Announce Type: new Abstract: The fundamental goal of agentic visual reasoning is to improve the success rate of multimodal large language models (MLLMs) on complex tasks, rather tha

arXiv:2510.05592v2 Announce Type: replace Abstract: Outcome-driven reinforcement learning has advanced reasoning in large language models (LLMs), but prevailing tool-augmented approaches train a singl