Test-time scaling (TTS) has demonstrated remarkable potential in enhancing the reasoning capabilities of Large Language Models (LLMs) and Large Vision-Language Models (LVLMs). However, its application has primarily been limited to domains such as mathematics and programming, owing to their reasoning-intensive nature and the ease of result verification. Its utility in other knowledge-intensive fields, such as medicine and general scientific research, remains underexplored. To bridge this gap and unlock the potential of TTS in broader domains, we propose Cross-Domain TTS, a novel framework that enables task-tailored scaling. This framework consists of two key components: a conformal prediction-based cold-start strategy and an information-gain-based dynamic reasoning adjustment. The CP-based cold-start strategy guides the model's initialization during test-time scaling based on conformal prediction theory, while the information-gain-based dynamic reasoning adjustment guides the model's reasoning progress through a progress vector according to the information gain of reasoning steps. We conducted experiments using LLMs and LVLMs on cross-domain benchmarks. Our results demonstrate that the proposed framework consistently improves performance across various domain-specific datasets. For instance, in the medical domain, it achieves an improvement of up to 17% in pass@1 accuracy while reducing inference latency and saving up to 30% in token consumption. Code is available at https://github.com/Yan0613/Cross-Domain-TTS.
论文检索
输入标题、作者或关键词，从 15,207 篇学术成果中精准定位
Graphical User Interface (GUI) Agents, powered by large language and vision-language models, hold promise for enabling end-to-end automation in digital environments. However, their progress is fundamentally constrained by the scarcity of scalable, high-quality trajectory data. Existing data collection strategies either rely on costly and inconsistent manual annotations or on synthetic generation methods that trade off between diversity and meaningful task coverage. To bridge this gap, we present GUI-ReWalk: a reasoning-enhanced, multi-stage framework for synthesizing realistic and diverse GUI trajectories. GUI-ReWalk begins with a stochastic exploration phase that emulates human trial-and-error behaviors, and progressively transitions into a reasoning-guided phase where inferred goals drive coherent and purposeful interactions. Moreover, it supports multi-stride task generation, enabling the construction of long-horizon workflows across multiple applications. By combining randomness for diversity with goal-aware reasoning for structure, GUI-ReWalk produces data that better reflects the intent-aware, adaptive nature of human-computer interaction. We further train Qwen2.5-VL-7B on the GUI-ReWalk dataset and evaluate it across multiple benchmarks, including Screenspot-Pro, OSWorld/OSWorld-G, UI-Vision, AndroidControl, and GUI-Odyssey. Results demonstrate that GUI-ReWalk enables superior coverage of diverse interaction flows, higher trajectory entropy, and more realistic user intent. These findings establish GUI-ReWalk as a scalable and data-efficient framework for advancing GUI agent research and enabling robust real-world automation.
In this paper, we revisit multimodal few-shot 3D point cloud semantic segmentation (FS-PCS), identifying a conflict in "Fuse-then-Refine" paradigms: the "Plasticity-Stability Dilemma." In addition, Contrastive Language-Image Pre-training (CLIP)'s inter-class confusion can result in semantic blindness. To address these issues, we present the Decoupled-experts Arbitration Few-Shot SegNet (DA-FSS), a model that effectively distinguishes between semantic and geometric paths and mutually regularizes their gradients to achieve better generalization. DA-FSS employs the same backbone and pre-trained text encoder as the baseline MultiModal Few-Shot SegNet (MM-FSS) to generate text embeddings, which can increase cost-free modalities' utilization rate and better leverage each modality's information space. To achieve this, we propose a Parallel Expert Refinement module to generate each modal correlation. We also propose a Stacked Arbitration Module (SAM) to perform convolutional fusion and arbitrate correlations for each modality pathway. The Parallel Experts decouple two paths: a Geometric Expert maintains plasticity, and a Semantic Expert ensures stability. They are coordinated via a Decoupled Alignment Module (DAM) that transfers knowledge without propagating confusion. Experiments on popular datasets (S3DIS, ScanNet) demonstrate the superiority of DA-FSS over MM-FSS. Meanwhile, geometric boundaries, completeness, and texture differentiation are all superior to the baseline. The code is available at: https://github.com/MoWenQAQ/DA-FSS/.
Recent efforts have successfully scaled plain Vision Transformers (ViTs) to unprecedented sizes, ranging from 6B to 22B parameters. However, hierarchical ViTs, which are inherently better suited for multi-scale representation and dense prediction tasks, have largely remained constrained to under 2B parameters. This scaling gap prevents the community from leveraging high-capacity, multi-scale foundation models for complex visual scenes. To bridge this gap, we propose EHV, an Efficient Hierarchical ViT architecture designed for massive scaling. Our dense models scale from 200M to 5B parameters, and we further introduce a Sparse Mixture-of-Experts (SMoE) variant, pushing the scale to an industry-leading 30B parameters. The training pipeline follows a two-stage pretraining strategy: (1) MAE Stage: self-supervised pretraining on ImageNet-21K using a Masked Autoencoder (MAE); and (2) Distillation Stage: distilling knowledge from multiple state-of-the-art foundation models on a 27M-image dataset. With only 6.7B active parameters, EHV-5B-MoE achieves an exceptional 89.0% linear accuracy on ImageNet-1K, outperforming much larger models like EVA-CLIP-18B and DINOv3-7B. These results demonstrate that EHV effectively learns high-quality, generalizable, and linearly separable features across image classification, video analysis, and dense prediction tasks.
Few-shot learning aims to recognize novel categories from limited labeled samples, where prototypes estimated from 1--5 supports per class are often unreliable. Semantic-based approaches alleviate this by introducing class-level priors, but they often ignore instance-level cues and rarely optimize queries under the inductive protocol. We propose PMCE, a Probabilistic framework that leverages Multi-granularity semantics with Caption-guided Enhancement for few-shot classification. On base classes, we build a knowledge bank with class-wise visual statistics and class-name embeddings. At test time, the semantic embedding of a novel class retrieves a few similar base classes whose visual priors are aggregated into a class-specific prior and combined with the support-based prototype via MAP estimation. Simultaneously, we train a lightweight enhancer on base classes that fuses frozen BLIP captions with visual features, and apply it to both supports and queries without using query-set statistics or novel labels. A simple caption-consistency regularizer further improves robustness to noisy captions. Experiments on four standard benchmarks with ResNet-12 and Swin-T backbones show that PMCE outperforms state-of-the-art few-shot baselines, achieving up to 7.71% gains over the strongest competitor on MiniImageNet in the challenging 1-shot setting. Our code is available at https://github.com/channa419/PMCE.
Event Causality Identification (ECI) is a crucial task in knowledge discovery that extracts structured causal relationships between annotated event mentions from unstructured text. However, existing approaches typically rely on extensive labeled data, which is scarce for specialized domains and topics. Although Large Language Models (LLMs) show strong promise for few-shot and zero-shot information extraction, they are prone to “causal hallucination,” generating unreliable and spurious causal links. To address these limitations, we propose LLM-SD (Large Language Model Self-Debate), a novel framework that formulates ECI as a structured debate among multiple identical instances of a single LLM. Causality is determined through the integration of adversarial evidence. The framework employs LLMs in distinct roles: an affirmative team argues for the existence of causality, a negative team argues against it, and an adjudication committee evaluates the evidence for determination. An evidence strength grading rule guides the quantification and integration of adversarial evidence. The automatic and LLM-driven verification finally produce a reasoned verdict for event causality. LLM-SD reduces spurious causal links resulting from causal hallucination in LLMs and identifies more long-distance causalities by promoting a balanced evaluation of arguments. Extensive experiments on three benchmark datasets demonstrate that LLM-SD achieves state-of-the-art performance in a zero-shot setting.
Ethical awareness is critical for robots operating in human environments, yet existing automated planning tools provide little support. Manually specifying ethical rules is labour-intensive and highly context-specific. We present Principles2Plan, an interactive research prototype demonstrating how a human and a Large Language Model (LLM) can collaborate to produce context-sensitive ethical rules and guide automated planning. A domain expert provides the planning domain, problem details, and relevant high-level principles such as beneficence and privacy. The system generates operationalisable ethical rules consistent with these principles, which the user can review, prioritise, and supply to a planner to produce ethically-informed plans. To our knowledge, no prior system supports users in generating principle-grounded rules for classical planning contexts. Principles2Plan showcases the potential of human-LLM collaboration for making ethical automated planning more practical and feasible.
Causal agents have emerged as promising tools for automating causal analysis based on user queries. However, existing causal agent systems are often limited to a single causal task, limiting their ability to handle complex queries. In addition, they accept only numerical data as input, preventing the integration of domain knowledge expressed in natural language. To overcome these limitations, we propose the OrcheCause agent, a causal agent leveraging textual knowledge for end-to-end causal inference. Specifically, OrcheCause is designed to orchestrate a sequence of interrelated causal tasks in response to user queries. Furthermore, OrcheCause supports diverse data types—numerical as well as textual data—by extracting cause-effect pairs from the relevant sources and incorporating them into causal discovery (CD), thereby improving the performance of CD. OrcheCause also introduces a metric-based hyperparameter optimization framework for CD when ground-truth graphs are not available.
Despite the rapid advancement of generative agents, their deployment in real world industry scenarios often encounters challenges due to a lack of domain-specific knowledge. To address this gap, we present KnowPilot: a Domain-Specific Knowledge Augmented Agent System. KnowPilot is an open-source framework that integrates task-specific priors, explicit knowledge, and experiential knowledge to enhance agent performance in specialized applications. It combines knowledge retrieval from structured repositories with a memory system capable of capturing expert experience through human–AI interaction.
AgentGraph: Trace-to-Graph Platform for Interactive Analysis and Robustness Testing in Agentic AI Systems
PDF ↗Modern Agentic AI systems plan, reason, and act across multiple steps, creating execution patterns that are difficult to interpret. Existing observability platforms track prompt I/O and operational metrics but require manual inspection of traces to reconstruct structure and reasoning. We present AgentGraph, which converts execution logs into interactive knowledge graphs and actionable insights. Nodes represent agents, tasks, tools, data inputs/outputs, and humans, while typed edges capture relations such as inputs consumed, tasks delegated or sequenced, tools required or used, outputs produced and delivered, and interventions from agents or humans. Each graph element links to its exact trace span, ensuring verifiability. Building on this representation, AgentGraph enables two analyses: qualitative trace-grounded failure detection and optimisation recommendations, and quantitative robustness evaluation via perturbation testing and causal attribution.
Capturing expertise and enabling efficient information retrieval are critical in the energy sector, where high staff turnover can lead to significant knowledge loss. Retrieval Augmented Generation (RAG) offers a solution by grounding Large Language Model (LLM) outputs in documented sources, but its effectiveness is limited by reliance on general-purpose embeddings. We present Wikatoni, an agentic AI system for energy engineering workflows that integrates a novel domain-specific embedding model. Wikatoni combines fine-tuned embeddings with agentic RAG, metadata filtering, and hybrid retrieval to improve document search, automated reporting, and workflow efficiency. Evaluation on internal enterprise offshore energy data shows that the domain-adapted embedding improves recall by 10%, and Wikatoni agentic RAG further increases answer accuracy by 14% compared to vanilla RAG with the base embedding model, achieving the best overall performance in context recall, faithfulness, and answer accuracy.
Multi-Agent Path Finding (MAPF) algorithms provide highly optimized solutions for coordinating multiple agents in shared environments, yet their outputs lack explainability to human stakeholders. Existing explanation approaches, such as visual trace segmentation or logic-based reasoning, remain fragmented. In this demo, we present OMEGA, an interactive explanation platform that generates Natural Language (NL) explanations using the novel Multi-Agent Planning Ontology (maPO). Our framework transforms raw MAPF planner execution logs into a semantic knowledge graph, enabling SPARQL-based explanations of collision events, replanning strategies, and efficiency trade-offs. A lightweight web interface allows users to query, visualize, and interpret planner decisions, thereby making MAPF solutions transparent and auditable. We conducted a user study that confirms the ontology-driven explanations are significantly clearer and more preferred than raw logs, underscoring the potential of semantic technologies for explainable multi-agent systems.
GeoProblem Factory: A Visual Interaction System for Solvable and Controllable Geometric Problem Generation by Leveraging Symbolic Deduction Engine
PDF ↗We propose a novel system, GeoProblem Factory, designed to effectively generate high-quality geometry problems for intelligent education. The system enables to efficiently produce batches of geometry problems for teachers and students, either to save time and manual effort or to support personalized learning. Generating geometry problems is particularly challenging, as it requires ensuring both solvability and controllability from a pedagogical perspective. To address these issues, we adopt a state-of-the-art pipeline method based on a symbolic deduction engine and develop a visual interaction demo. This demo allows users to easily refine the generated problems through visual operations. It provides two modes for inputting controllable information: specifying knowledge points or supplying a reference problem. Moreover, the system can automatically generate a preliminary geometric diagram corresponding to each problem for further refinement. Through human–machine interaction, the system can more efficiently produce high-quality geometry problems than ever.
Large language models (LLMs) often produce factually inaccurate content, or hallucinations, which undermines their reliability. Existing factuality evaluation systems usually rely on a single predefined fact source, making them task-specific and hard to extend. We present UFO, a unified framework for factuality evaluation that supports multiple plug-and-play fact sources. UFO integrates human-written evidence, web search results, and LLM knowledge within a single evaluation pipeline, and allows users to flexibly select, reorder, and even define customized sources. The system is accessible through both a Python interface and a web-based demo, offering interactive claim-level verification and visualization. Experiments show that UFO system achieves moderate consistency with human annotations. Overall, UFO serves as a transparent and extensible platform for benchmarking fact sources, comparing LLMs, and enabling real-world fact-checking applications across diverse domains.
Language models are powerful artifacts, yet their factual knowledge is still poorly understood, and inaccessible to ad-hoc browsing and scalable statistical analysis. This demonstration introduces GPTKB v1.5, a densely interlinked 100-million-triple knowledge base (KB) built for $14,000 from GPT-4.1, using the GPTKB methodology for massive-recursive LLM knowledge materialization. This demo focuses on three use cases: (1) link-traversal-based LLM knowledge exploration, (2) SPARQL-based structured LLM knowledge querying, (3) comparative exploration of the strengths and weaknesses of LLM knowledge. Massive-recursive LLM knowledge materialization is a groundbreaking opportunity both for the systematic analysis of LLM knowledge, as well as for automated KB construction.
Retrieval-Augmented Generation (RAG) has become the standard approach for integrating domain knowledge into Large Language Models (LLMs). However, fair comparison of RAG pipelines remains difficult: data preparation is often ad hoc, subsampling methods are opaque, parameters vary across implementations, and evaluation is fragmented. We present In-Situ Eval, a unified and reproducible framework that operationalizes the full RAG pipeline with configurable subsampling strategies and both RAG-specific and generic evaluation metrics. The platform supports two execution modes: an offline Dataset mode for evaluating precomputed outputs, and a live Retrieval mode for benchmarking RAG variants with state-of-the-art LLMs. Users can flexibly select datasets, retrieval techniques, models, and metrics, enabling side-by-side comparisons, ablations, and targeted analyses. This holistic approach reduces computational costs, clarifies the impact of subsampling techniques, and provides actionable insights for real-world deployments. By facilitating transparent, customizable, and interactive benchmarking, In-Situ Eval empowers both researchers and practitioners to make informed decisions in adapting RAG pipelines to domain-specific needs.
We present a Proactive Monitoring Agent designed for large-scale customer data platforms, such as Adobe Experience Platform (AEP), to predict and prevent workflow disruptions before they impact business operations. Unlike existing reactive solutions that assist engineers only after failures occur, our agent anticipates potential failures across multiple workflow stages, explains its predictions in natural language, and interacts with customer support engineers through a conversational interface. The system integrates a machine learning-based Prediction Module, Knowledge Graph APIs for contextual data access, and a Query Processor that powers an interactive Q&A experience, enabling timely and actionable insights to minimize operational risks and maximize business continuity.
Scientific research articles, typically distributed in PDF format, contain valuable knowledge but remain challenging to convert into structured datasets due to fragmented workflows that separate parsing, annotation, and visualization. Existing annotation platforms operate on plain text, which requires an additional PDF-to-text conversion step before annotation, while PDF parsing tools lack automated annotation suggestions. To bridge this gap, we introduce Docora, a system that unifies PDF parsing, automated annotation assistance, and multi-view visualization into a single interactive platform. Docora enables researchers to configure entity and relation schemas for any domain, automatically generates initial annotations using rule-based, model-based, or LLM-based extractors, and provides synchronized visualizations across PDF, text, and graph views. Users can refine annotations directly on the PDF canvas, ensuring consistency between document layout and structured representations. The system’s source code is publicly available to facilitate further research and development.
Knowledge Graphs (KGs) provide structured, verifiable representations that ground facts and supply large language models (LLMs) with reliable real-world information. Building high-quality KGs from open-domain text remains difficult due to redundancy, inconsistency, and lack of ontology grounding. We present Wikontic, a pipeline that extracts triples from text with LLMs and refines them through ontology-based typing, schema validation, and entity deduplication, yielding compact and coherent graphs. Unlike prior frameworks that lack ontology grounding or perform only partial deduplication, Wikontic uniquely integrates entity canonicalization, alias tracking, and automatic enforcement of Wikidata’s ontology, enabling robust schema-aware construction without manual schema design. Its web interface lets users upload text, visualize graphs, and perform multi-hop question answering. By combining LLM flexibility with Wikidata’s ontological rigor, Wikontic transforms ambiguous text into structured, interpretable, and actionable knowledge.
Supporting children with Autism Spectrum Disorder (ASD) requires highly individualized knowledge. However, critical information is often dispersed across documents such as Individualized Education Plans (IEPs), diagnostic assessments, and caregiver notes. Thus, we propose SHARE (Synthesizing Heterogeneous Autism-support Records into Evidence-based Recommendations), a framework that combines diverse autism-related documents into a concise, actionable set of recommendations for caregivers of children with ASD. Feedback is generated using OpenAI’s large language model API, grounded in user-provided evidence with optional web-based extensions for missing details, and citation-linked. After caregivers attempt and then rate recommendations, SHARE uses a Bayesian bandit algorithm with Upper Confidence Bound (UCB) re-ranking to refine future advice. While previous work mostly focuses on drafting static goals, SHARE additionally combines LLM-generated recommendations, caregiver feedback, and interpretable ranking into a pipeline that can adapt over time.