To address key bottlenecks in automated multiple-choice question (MCQ) generation—namely, shallow cognitive modeling, distractors lacking discipline-specific misconceptions, and poor scalability—this paper proposes a hierarchical concept graph–guided LLM collaborative generation framework. The method integrates structured physics concept graph modeling, domain-adaptive retrieval-augmented generation (RAG), misconception-driven prompt engineering, and an automated quality validation pipeline. It is the first to enable cognitively layered MCQ generation with misconception-embedded distractors. Experimental results demonstrate significant improvements: expert evaluation pass rate reaches 75.2% (vs. baseline 37.0%), and student random-guessing rate drops to 28.05% (vs. baseline 37.10%). These gains markedly enhance diagnostic validity and feedback efficiency of assessments, establishing a novel paradigm for intelligent educational assessment.

It remains unclear whether large language models (LLMs) possess neurons dedicated exclusively to abstract semantic relation representation—decoupled from specific entity instantiations. Method: Leveraging the Llama-2 family, we propose a statistics-driven neuron localization method, integrating selective ablation with multi-relation and cross-lingual comparative experiments. Contribution/Results: We provide the first empirical evidence of “relation-specific neurons.” These neurons exhibit three key properties: (1) cumulative activation—response magnitude increases monotonically with the number of relation instances; (2) strong generalization across distinct relations and languages; and (3) targeted ablation significantly impairs generation accuracy for the target relation while unexpectedly improving performance on non-target relations (positive transfer). This demonstrates a modular, relation-centric knowledge organization in LLMs, advancing interpretability research and enabling more controllable, fact-aware reasoning.

This work addresses the challenge of ensuring atomicity and consistency in offline-first peer-to-peer systems characterized by sparse connectivity and partial replication. To this end, it introduces— for the first time—the IntersectionAtomicity and IntersectionCC models, which formally capture the constraints on atomicity and consistency under transactional semantics in such environments. By integrating principles from distributed transaction theory with access control mechanisms, the paper develops a reasoning framework tailored to offline-first collaborative systems. Building upon this foundation, it derives practical correctness guarantees that directly inform system design, thereby significantly enhancing the reliability and correctness of applications operating under these challenging conditions.

This work addresses the sample complexity of Sample Average Approximation (SAA) for convex and strongly convex stochastic programming (SP) under standard SP assumptions—without requiring uniform Lipschitz continuity. Methodologically, it integrates convex analysis, stochastic optimization, and functional inequalities to bypass entropy-based arguments. The key contribution is the first tight, metric-entropy-free sample complexity bounds: $O(1/varepsilon^2)$ for convex SP and $O(1/varepsilon)$ for strongly convex SP—improving upon state-of-the-art bounds by a factor of $O(d)$ by eliminating dependence on covering numbers or Rademacher complexity. Theoretically, it reveals that SAA achieves nearly identical sample efficiency as stochastic mirror descent, bridging a long-standing gap in theoretical understanding. Numerical experiments validate the tightness of the bounds and demonstrate that SAA exhibits provably superior practical performance over stochastic mirror descent in non-Lipschitz settings.