Inference-Time Scaling of Verification: Self-Evolving Deep Research Agents via Test-Time Rubric-Guided Verification

Recent advances in Deep Research Agents (DRAs) are transforming automated knowledge discovery and problem-solving. While the majority of existing efforts focus on enhancing policy capabilities via post-training, we propose an alternative paradigm: self-evolving the agent's ability by iteratively verifying the policy model's outputs, guided by meticulously crafted rubrics. This approach gives rise to the inference-time scaling of verification, wherein an agent self-improves by evaluating its generated answers to produce iterative feedback and refinements. We derive the rubrics based on an automatically constructed DRA Failure Taxonomy, which systematically classifies agent failures into five major categories and thirteen sub-categories. We present DeepVerifier, a rubrics-based outcome reward verifier that leverages the asymmetry of verification and outperforms vanilla agent-as-judge and LLM judge baselines by 12%-48% in meta-evaluation F1 score. To enable practical self-evolution, DeepVerifier integrates as a plug-and-play module during test-time inference. The verifier produces detailed rubric-based feedback, which is fed back to the agent for iterative bootstrapping, refining responses without additional training. This test-time scaling delivers 8%-11% accuracy gains on challenging subsets of GAIA and XBench-DeepResearch when powered by capable closed-source LLMs. Finally, to support open-source advancement, we release DeepVerifier-4K, a curated supervised fine-tuning dataset of 4,646 high-quality agent steps focused on DRA verification. These examples emphasize reflection and self-critique, enabling open models to develop robust verification capabilities.

Why Does the LLM Stop Computing: An Empirical Study of User-Reported Failures in Open-Source LLMs

The democratization of open-source Large Language Models (LLMs) allows users to fine-tune and deploy models on local infrastructure but exposes them to a First Mile deployment landscape. Unlike black-box API consumption, the reliability of user-managed orchestration remains a critical blind spot. To bridge this gap, we conduct the first large-scale empirical study of 705 real-world failures from the open-source DeepSeek, Llama, and Qwen ecosystems. Our analysis reveals a paradigm shift: white-box orchestration relocates the reliability bottleneck from model algorithmic defects to the systemic fragility of the deployment stack. We identify three key phenomena: (1) Diagnostic Divergence: runtime crashes distinctively signal infrastructure friction, whereas incorrect functionality serves as a signature for internal tokenizer defects. (2) Systemic Homogeneity: Root causes converge across divergent series, confirming reliability barriers are inherent to the shared ecosystem rather than specific architectures. (3) Lifecycle Escalation: Barriers escalate from intrinsic configuration struggles during fine-tuning to compounded environmental incompatibilities during inference. Supported by our publicly available dataset, these insights provide actionable guidance for enhancing the reliability of the LLM landscape.

From Laboratory to Real-World Applications: Benchmarking Agentic Code Reasoning at the Repository Level

As large language models (LLMs) evolve into autonomous agents, evaluating repository-level reasoning, the ability to maintain logical consistency across massive, real-world, interdependent file systems, has become critical. Current benchmarks typically fluctuate between isolated code snippets and black-box evaluations. We present RepoReason, a white-box diagnostic benchmark centered on abductive assertion verification. To eliminate memorization while preserving authentic logical depth, we implement an execution-driven mutation framework that utilizes the environment as a semantic oracle to regenerate ground-truth states. Furthermore, we establish a fine-grained diagnostic system using dynamic program slicing, quantifying reasoning via three orthogonal metrics: $ESV$ (reading load), $MCL$ (simulation depth), and $DFI$ (integration width). Comprehensive evaluations of frontier models (e.g., Claude-4.5-Sonnet, DeepSeek-v3.1-Terminus) reveal a prevalent aggregation deficit, where integration width serves as the primary cognitive bottleneck. Our findings provide granular white-box insights for optimizing the next generation of agentic software engineering.

Next Edit Prediction: Learning to Predict Code Edits from Context and Interaction History

The rapid advancement of large language models (LLMs) has led to the widespread adoption of AI-powered coding assistants integrated into a development environment. On one hand, low-latency code completion offers completion suggestions but is fundamentally constrained to the cursor's current position. On the other hand, chat-based editing can perform complex modifications, yet forces developers to stop their work, describe the intent in natural language, which causes a context-switch away from the code. This creates a suboptimal user experience, as neither paradigm proactively predicts the developer's next edit in a sequence of related edits. To bridge this gap and provide the seamless code edit suggestion, we introduce the task of Next Edit Prediction, a novel task designed to infer developer intent from recent interaction history to predict both the location and content of the subsequent edit. Specifically, we curate a high-quality supervised fine-tuning dataset and an evaluation benchmark for the Next Edit Prediction task. Then, we conduct supervised fine-tuning on a series of models and performed a comprehensive evaluation of both the fine-tuned models and other baseline models, yielding several novel findings. This work lays the foundation for a new interaction paradigm that proactively collaborate with developers by anticipating their following action, rather than merely reacting to explicit instructions.

ScreenCoder: Advancing Visual-to-Code Generation for Front-End Automation via Modular Multimodal Agents

Automating the transformation of user interface (UI) designs into front-end code holds significant promise for accelerating software development and democratizing design workflows. While recent large language models (LLMs) have demonstrated progress in text-to-code generation, many existing approaches rely solely on natural language prompts, limiting their effectiveness in capturing spatial layout and visual design intent. In contrast, UI development in practice is inherently multimodal, often starting from visual sketches or mockups. To address this gap, we introduce a modular multi-agent framework that performs UI-to-code generation in three interpretable stages: grounding, planning, and generation. The grounding agent uses a vision-language model to detect and label UI components, the planning agent constructs a hierarchical layout using front-end engineering priors, and the generation agent produces HTML/CSS code via adaptive prompt-based synthesis. This design improves robustness, interpretability, and fidelity over end-to-end black-box methods. Furthermore, we extend the framework into a scalable data engine that automatically produces large-scale image-code pairs. Using these synthetic examples, we fine-tune and reinforce an open-source VLM, yielding notable gains in UI understanding and code quality. Extensive experiments demonstrate that our approach achieves state-of-the-art performance in layout accuracy, structural coherence, and code correctness. Our code is made publicly available at https://github.com/leigest519/ScreenCoder.

Runtime Failure Hunting for Physics Engine Based Software Systems: How Far Can We Go?

Physics Engines (PEs) are fundamental software frameworks that simulate physical interactions in applications ranging from entertainment to safety-critical systems. Despite their importance, PEs suffer from physics failures, deviations from expected physical behaviors that can compromise software reliability, degrade user experience, and potentially cause critical failures in autonomous vehicles or medical robotics. Current testing approaches for PE-based software are inadequate, typically requiring white-box access and focusing on crash detection rather than semantically complex physics failures. This paper presents the first large-scale empirical study characterizing physics failures in PE-based software. We investigate three research questions addressing the manifestations of physics failures, the effectiveness of detection techniques, and developer perceptions of current detection practices. Our contributions include: (1) a taxonomy of physics failure manifestations; (2) a comprehensive evaluation of detection methods including deep learning, prompt-based techniques, and large multimodal models; and (3) actionable insights from developer experiences for improving detection approaches. To support future research, we release PhysiXFails, code, and other materials at https://sites.google.com/view/physics-failure-detection.

3D Software Synthesis Guided by Constraint-Expressive Intermediate Representation

Graphical user interface (UI) software has undergone a fundamental transformation from traditional two-dimensional (2D) desktop/web/mobile interfaces to spatial three-dimensional (3D) environments. While existing work has made remarkable success in automated 2D software generation, such as HTML/CSS and mobile app interface code synthesis, the generation of 3D software still remains under-explored. Current methods for 3D software generation usually generate the 3D environments as a whole and cannot modify or control specific elements in the software. Furthermore, these methods struggle to handle the complex spatial and semantic constraints inherent in the real world. To address the challenges, we present Scenethesis, a novel requirement-sensitive 3D software synthesis approach that maintains formal traceability between user specifications and generated 3D software. Scenethesis is built upon ScenethesisLang, a domain-specific language that serves as a granular constraint-aware intermediate representation (IR) to bridge natural language requirements and executable 3D software. It serves both as a comprehensive scene description language enabling fine-grained modification of 3D software elements and as a formal constraint-expressive specification language capable of expressing complex spatial constraints. By decomposing 3D software synthesis into stages operating on ScenethesisLang, Scenethesis enables independent verification, targeted modification, and systematic constraint satisfaction. Our evaluation demonstrates that Scenethesis accurately captures over 80% of user requirements and satisfies more than 90% of hard constraints while handling over 100 constraints simultaneously. Furthermore, Scenethesis achieves a 42.8% improvement in BLIP-2 visual evaluation scores compared to the state-of-the-art method.

Entropy-Memorization Law: Evaluating Memorization Difficulty of Data in LLMs

Large Language Models (LLMs) are known to memorize portions of their training data, sometimes reproducing content verbatim when prompted appropriately. In this work, we investigate a fundamental yet under-explored question in the domain of memorization: How to characterize memorization difficulty of training data in LLMs? Through empirical experiments on OLMo, a family of open models, we present the Entropy-Memorization Law. It suggests that data entropy is linearly correlated with memorization score. Moreover, in a case study of memorizing highly randomized strings, or "gibberish", we observe that such sequences, despite their apparent randomness, exhibit unexpectedly low empirical entropy compared to the broader training corpus. Adopting the same strategy to discover Entropy-Memorization Law, we derive a simple yet effective approach to distinguish training and testing data, enabling Dataset Inference (DI).

SlideCoder: Layout-aware RAG-enhanced Hierarchical Slide Generation from Design

Manual slide creation is labor-intensive and requires expert prior knowledge. Existing natural language-based LLM generation methods struggle to capture the visual and structural nuances of slide designs. To address this, we formalize the Reference Image to Slide Generation task and propose Slide2Code, the first benchmark with difficulty-tiered samples based on a novel Slide Complexity Metric. We introduce SlideCoder, a layout-aware, retrieval-augmented framework for generating editable slides from reference images. SlideCoder integrates a Color Gradient-based Segmentation algorithm and a Hierarchical Retrieval-Augmented Generation method to decompose complex tasks and enhance code generation. We also release SlideMaster, a 7B open-source model fine-tuned with improved reverse-engineered data. Experiments show that SlideCoder outperforms state-of-the-art baselines by up to 40.5 points, demonstrating strong performance across layout fidelity, execution accuracy, and visual consistency. Our code is available at https://github.com/vinsontang1/SlideCoder.

DesignBench: A Comprehensive Benchmark for MLLM-based Front-end Code Generation

Multimodal Large Language Models (MLLMs) have demonstrated remarkable capabilities in automated front-end engineering, e.g., generating UI code from visual designs. However, existing front-end UI code generation benchmarks have the following limitations: (1) While framework-based development becomes predominant in modern front-end programming, current benchmarks fail to incorporate mainstream development frameworks. (2) Existing evaluations focus solely on the UI code generation task, whereas practical UI development involves several iterations, including refining editing, and repairing issues. (3) Current benchmarks employ unidimensional evaluation, lacking investigation into influencing factors like task difficulty, input context variations, and in-depth code-level analysis. To bridge these gaps, we introduce DesignBench, a multi-framework, multi-task evaluation benchmark for assessing MLLMs' capabilities in automated front-end engineering. DesignBench encompasses three widely-used UI frameworks (React, Vue, and Angular) alongside vanilla HTML/CSS, and evaluates on three essential front-end tasks (generation, edit, and repair) in real-world development workflows. DesignBench contains 900 webpage samples spanning over 11 topics, 9 edit types, and 6 issue categories, enabling detailed analysis of MLLM performance across multiple dimensions. Our systematic evaluation reveals critical insights into MLLMs' framework-specific limitations, task-related bottlenecks, and performance variations under different conditions, providing guidance for future research in automated front-end development. Our code and data are available at https://github.com/WebPAI/DesignBench.