Large Language Models (LLMs) have undergone continuous advancement, achieving significant breakthroughs in both inference speed and output quality, while increasingly gaining the capability to select and invoke external tools. A growing number of LLM-based agents have emerged--capable not only of engaging in multi-turn dialogues or solving International Mathematical Olympiad (IMO) level problems, but also of autonomously planning actions, making decisions, and interacting with external APIs, databases, and tools when faced with complex tasks. However, disparate databases, web services, and applications remain largely siloed, posing substantial engineering complexity for developers due to the lack of seamless integration and extensibility. To address this challenge, the Model Context Protocol (MCP) [1] has been introduced as a standardized interface for connecting LLMs with external data sources. MCP significantly reduces integration overhead and establishes a secure, trusted communication channel between MCP clients and servers, thereby fulfilling the scalability and interoperability requirements of AI-powered services.

ABSTRACT We proposed Static and Dynamic--two zero-shot logits-layer debiasing methods. Dynamic reduces bias by up to 70% with minimal fluency loss. Logits intervention outperforms hidden-layer approaches. We show semantic-aware logits intervention is stable and effective for debiasing aligned LLMs. Index T erms-- LLM Alignment,Debiasing 1. INTRODUCTION The rapid advancement of Large Language Models (LLMs) has revolutionized natural language processing, but their growing complexity raises critical concerns about trustworthiness [1].

Large language models (LLMs) have recently enabled coding agents capable of generating, executing, and revising visualization code. However, existing models often fail in practical workflows due to limited language coverage, unreliable execution, and lack of iterative correction mechanisms. Progress has been constrained by narrow datasets and benchmarks that emphasize single-round generation and single-language tasks. To address these challenges, we introduce three complementary resources for advancing visualization coding agents. VisCode-Multi-679K is a large-scale, supervised dataset containing 679K validated and executable visualization samples with multi-turn correction dialogues across 12 programming languages. VisPlotBench is a benchmark for systematic evaluation, featuring executable tasks, rendered outputs, and protocols for both initial generation and multi-round self-debug. Finally, we present VisCoder2, a family of multi-language visualization models trained on VisCode-Multi-679K. Experiments show that VisCoder2 significantly outperforms strong open-source baselines and approaches the performance of proprietary models like GPT-4.1, with further gains from iterative self-debug, reaching 82.4% overall execution pass rate at the 32B scale, particularly in symbolic or compiler-dependent languages.

Large language models (LLMs) have recently demonstrated impressive multimodal reasoning capabilities, yet their understanding of purely numerical time-series signals remains limited. Existing approaches mainly focus on forecasting or trend description, without uncovering the latent events that drive numerical changes or explaining the reasoning process behind them. In this work, we introduce the task of number-to-event reasoning and decoding, which aims to infer interpretable structured events from numerical inputs, even when current text is unavailable. To address the data scarcity and semantic alignment challenges, we propose a reasoning-aware framework that integrates an agent-guided event extractor (AGE), a marked multivariate Hawkes-based synthetic generator (EveDTS), and a two-stage fine-tuning pipeline combining a time-series encoder with a structured decoder. Our model explicitly reasons over numerical changes, generates intermediate explanations, and outputs structured event hypotheses. Experiments on multi-domain datasets show that our method substantially outperforms strong LLM baselines in event-level precision and recall. These results suggest a new direction for bridging quantitative reasoning and semantic understanding, enabling LLMs to explain and predict events directly from numerical dynamics.

Building robust and general reasoning ability is a central goal in the development of large language models (LLMs). Recent efforts increasingly turn to code as a rich training source, given its inherent logical structure and diverse reasoning paradigms such as divide-and-conquer, topological ordering, and enumeration. However, reasoning in code is often expressed implicitly and entangled with syntactic or implementation noise, making direct training on raw code suboptimal.To address this, we introduce TracePile, a large-scale corpus of 2.6 million samples that transforms code execution into explicit, step-by-step chain-of-thought-style rationales, which we call Chain of Execution (CoE). The corpus spans domains including mathematics, classical algorithms and algorithmic competition, and is enriched with variable-tracing questions and code rewritings to enhance logical granularity and code diversity. We evaluate TracePile using three training setups: continue-pretraining, instruction tuning after pretraining, and two-stage finetuning. Experiments across four base models (LLaMA 3, LLaMA 3.1, Qwen-2.5, and Qwen-2.5 Coder) and 20 benchmarks covering math, code, logic, and algorithms demonstrate consistent improvements. Notably, TracePile boosts LLaMA3.1-8B by 7.1\% on average across nine math datasets and delivers clear gains on LiveCodeBench, CRUX, and MMLU under two-stage fine-tuning.

Xynapse Traces is an experimental publishing imprint created via a fusion of human and algorithmic methods using a configuration-driven architecture and a multi-model AI integration framework. The system achieved a remarkable 90% reduction in time-to-market (from a typical 6-12 months to just 2-4 weeks), with 80% cost reduction compared to traditional imprint development, while publishing 52 books in its first year and maintaining exceptional quality metrics, including 99% citation accuracy and 100% validation success after initial corrections. Key technical innovations include a continuous ideation pipeline with tournament-style evaluation, a novel codex design for transcriptive meditation practice, comprehensive automation spanning from ideation through production and distribution, and publisher personas that define and guide the imprint's mission. The system also integrates automated verification with human oversight, ensuring that gains in speed do not compromise publishing standards. This effort has significant implications for the future of book publishing, suggesting new paradigms for human-AI collaboration that democratize access to sophisticated publishing capabilities and make previously unviable niche markets accessible.

Social media user-generated content (UGC) provides real-time, self-reported indicators of mental health conditions such as depression, offering a valuable source for predictive analytics. While prior studies integrate medical knowledge to improve prediction accuracy, they overlook the opportunity to simultaneously expand such knowledge through predictive processes. We develop a Closed-Loop Large Language Model (LLM)-Knowledge Graph framework that integrates prediction and knowledge expansion in an iterative learning cycle. In the knowledge-aware depression detection phase, the LLM jointly performs depression detection and entity extraction, while the knowledge graph represents and weights these entities to refine prediction performance. In the knowledge refinement and expansion phase, new entities, relationships, and entity types extracted by the LLM are incorporated into the knowledge graph under expert supervision, enabling continual knowledge evolution. Using large-scale UGC, the framework enhances both predictive accuracy and medical understanding. Expert evaluations confirmed the discovery of clinically meaningful symptoms, comorbidities, and social triggers complementary to existing literature. We conceptualize and operationalize prediction-through-learning and learning-through-prediction as mutually reinforcing processes, advancing both methodological and theoretical understanding in predictive analytics. The framework demonstrates the co-evolution of computational models and domain knowledge, offering a foundation for adaptive, data-driven knowledge systems applicable to other dynamic risk monitoring contexts.

Confidence in LLMs is a useful indicator of model uncertainty and answer reliability. Existing work mainly focused on single-turn scenarios, while research on confidence in complex multi-turn interactions is limited. In this paper, we investigate whether LLM-based search agents have the ability to communicate their own confidence through verbalized confidence scores after long sequences of actions, a significantly more challenging task compared to outputting confidence in a single interaction. Experimenting on open-source agentic models, we first find that models exhibit much higher task accuracy at high confidence while having near-zero accuracy when confidence is low. Based on this observation, we propose Test-Time Scaling (TTS) methods that use confidence scores to determine answer quality, encourage the model to try again until reaching a satisfactory confidence level. Results show that our proposed methods significantly reduce token consumption while demonstrating competitive performance compared to baseline fixed budget TTS methods.

AI-based code generation is increasingly prevalent, with GitHub Copilot estimated to generate 46% of the code on GitHub. Accurately evaluating how well generated code aligns with developer intent remains a critical challenge. Traditional evaluation methods, such as unit tests, are often unscalable and costly. Syntactic similarity metrics (e.g., BLEU, ROUGE) fail to capture code functionality, and metrics like CodeBERTScore require reference code, which is not always available. To address the gap in reference-free evaluation, with few alternatives such as ICE-Score, this paper introduces MATCH, a novel reference-free metric. MATCH uses Contrastive Learning to generate meaningful embeddings for code and natural language task descriptions, enabling similarity scoring that reflects how well generated code implements the task. We show that MATCH achieves stronger correlations with functional correctness and human preference than existing metrics across multiple programming languages.

Large language models (LLMs) have demonstrated promising performance in generating diagnostic conclusions from imaging findings, thereby supporting radiology reporting, trainee education, and quality control. However, systematic guidance on how to optimize prompt design across different clinical contexts remains underexplored. Moreover, a comprehensive and standardized framework for assessing the trustworthiness of LLM-generated radiology reports is yet to be established. This study aims to enhance the trustworthiness of LLM-generated liver MRI reports by introducing a Multi-Dimensional Credibility Assessment (MDCA) framework and providing guidance on institution-specific prompt optimization. The proposed framework is applied to evaluate and compare the performance of several advanced LLMs, including Kimi-K2-Instruct-0905, Qwen3-235B-A22B-Instruct-2507, DeepSeek-V3, and ByteDance-Seed-OSS-36B-Instruct, using the SiliconFlow platform.