mlflow

The widespread adoption of open-source machine learning (ML) datasets and models has intensified risks including data poisoning, supply-chain attacks, and regulatory non-compliance. Method: This paper proposes the first verifiable, end-to-end ML provenance framework integrating Trusted Execution Environments (TEEs) and transparent logging—built upon SPDX/SLSA standards and leveraging Intel SGX, hash-chain-based immutable logging, and zero-knowledge proofs. Contribution/Results: The framework enables provable artifact authenticity, auditable end-to-end lineage, and co-guaranteed confidentiality and integrity—without compromising intellectual property rights over data or models. Evaluated on two real-world ML pipelines, it achieves 100% metadata tampering detection, full verifiable traceability from training to deployment, and negligible runtime overhead—demonstrating practical viability for secure, compliant ML operations.

To address the challenges of poor interoperability between machine learning (ML) models and scientific computing models, as well as system instability under high concurrency in drug discovery, this paper proposes a model management platform integrating large language model (LLM) agents and generative AI. Methodologically, we design a model gateway built upon MLOps principles and microservice architecture, supporting dynamic consensus-based modeling, model registration/retrieval, asynchronous execution, and result callbacks—orchestrated autonomously by an LLM agent for intelligent model governance. Our key contribution is the first introduction of a dynamic consensus mechanism into collaborative model scheduling, augmented by generative AI to enhance meta-information understanding and operational decision-making. Experimental evaluation demonstrates zero failure rate under sustained load from over 10,000 concurrent clients, significantly improving scheduling intelligence and system scalability. The platform establishes foundational infrastructure for AI-driven drug development.

This study addresses the widespread neglect of licensing terms and regulatory compliance in the deployment of machine learning models within open-source software, particularly in safety-critical contexts where associated risks are pronounced. The authors present the first systematic investigation of ML usage across 173 open-source projects on GitHub spanning 16 application domains. Through code inspection and contextual analysis, they evaluate each model’s role in decision-making, the presence of risk-mitigation strategies, and adherence to licensing requirements. The findings reveal that certain projects employ ML for high-stakes decisions without complying with applicable license conditions and often lack essential post-processing safeguards. This work uncovers critical compliance blind spots in the open-source ecosystem and provides an empirical foundation for developing compliance guidelines and automated detection tools.

Machine learning (ML) assets—including models, datasets, and metadata—suffer from fragmented documentation, isolated storage, inconsistent licensing, and inadequate discovery mechanisms, severely hindering reuse and management efficiency. To address these challenges, this paper introduces the first unified, lifecycle-aware classification system and systematic management framework for ML assets. It integrates asset cataloging, structured metadata modeling, lineage tracking, and semantic retrieval to explicitly tackle three core system-level challenges: scalability, traceability, and cross-domain unified indexing. The framework is realized through an open-source toolchain and validated via real-world system demonstrations, enabling license-aware, cross-domain asset discovery and compliant reuse. Our contribution provides researchers and practitioners with a deployable management paradigm and practical tools, advancing ML asset governance from ad hoc, experience-driven practices toward rigorous, engineering-based stewardship.

This study identifies critical issues—namely, inconsistent naming conventions, absent semantic versioning, and opaque documentation—in the release practices of 52,227 open-source pre-trained language models (PTLMs) on Hugging Face, undermining version reliability and experimental reproducibility. Employing a mixed empirical methodology—including large-scale web crawling, naming-pattern clustering, version-change provenance tracing, foundational model lineage reconstruction, and documentation completeness assessment—the authors reveal that PTLM version identifiers (e.g., v1.2) are largely arbitrary; 40.87% of weight updates lack explicit representation in version names or documentation. They further identify 148 distinct naming practices and reconstruct 299 foundational model derivation lineages. Based on these findings, the work proposes the first semantic versioning framework specifically designed for PTLMs, aiming to standardize model releases and enhance reproducibility across the NLP research community.

This study addresses the unclear usage patterns and functional demands surrounding current MLOps frameworks in open-source projects, which hinder their effective evolution. For the first time, it systematically links real-world framework adoption with user enhancement requests by analyzing GitHub dependencies, API invocations, and issue reports across eight prominent MLOps frameworks, employing qualitative coding and thematic mapping. The findings reveal that developers prefer customized integrations over out-of-the-box solutions, and that these frameworks are seldom directly embedded in GitHub Workflows, instead being primarily applied to core machine learning phases and infrastructure governance. Users most frequently request enhancements to core functionality, greater API exposure, and improved CI/CD integration, while increasingly adopting multiple frameworks in tandem.

This work addresses the inefficiencies in large-scale recommendation systems caused by maintaining separate models for different scenarios and objectives, which hinders development velocity and delays technology adoption. To overcome this, the authors propose the Standardized Model Template (SMT) framework, which leverages composable, standardized machine learning components to enable “design once, deploy everywhere,” uniformly accommodating diverse data distributions and optimization objectives. By decoupling model architecture from scenario-specific configurations, SMT reduces the complexity of technology deployment from O(n·2ᵏ) to O(n+k), breaking away from the conventional “one objective, one model” paradigm. Empirical evaluation on Meta’s ad ranking system demonstrates that SMT improves average cross-entropy by 0.63%, reduces engineering time per model iteration by 92%, and increases the throughput of technology-model pair adoption by 6.3×.

This study addresses the significant gap between the static capabilities of open-source large language models and their real-world performance in hosted API services, a discrepancy exacerbated by insufficient understanding of service-layer heterogeneity and dynamics. Leveraging multidimensional data collected in Q4 2025 from the AI Ping platform—including request logs, metadata, compatibility probes, price snapshots, and latency measurements—the work uncovers three key patterns: demand concentration inertia, supply-demand misalignment, and task-conditioned routing. Building on these insights, the paper reframes model deployment as a constrained statistical decision problem and demonstrates that intelligent routing reduces inference costs for Qwen3-32B by 37.8% and increases throughput for DeepSeek-V3.2 by approximately 90%.

This work addresses reproducibility challenges in collaborative machine learning, which often stem not merely from missing artifacts but from team misalignments in interpreting prior work, inconsistent component evolution, and difficulties in reconstructing experimental intent. To tackle these issues, the authors propose a novel two-layer socio-technical architecture that integrates interactive support into reproducibility frameworks for the first time. The lower layer leverages a data-centric ML management system with full lifecycle provenance tracking, while the upper layer employs an AI-mediated semantic interface to enable structured collaboration, explanatory discourse, and consensus building. Deployed over 19 months in a clinical research setting, the system effectively identified and mitigated persistent interaction breakdowns, substantially enhancing shared team understanding and experimental reproducibility.