CPL: A Compact C-like Systems Language with Explicit Low-Level Control
该论文介绍了一种紧凑的类C系统语言CPL，通过简化语法并结合新语言特性，旨在探索其在编译器实验中的便捷性，并展示了与主流C编译器相当的性能。
0 citationsRead paper
Selected work
Representative Papers
Which Values Do LLMs Confuse? A Schwartz-Based Recognition Study
This study addresses the limited reliability of large language models in identifying Schwartz’s ten basic human values, particularly their tendency to confuse adjacent value dimensions. The authors construct a balanced dataset comprising 1,000 Russian-language contextual prompts grounded in Schwartz’s theory and conduct the first systematic analysis of directional confusion patterns across 21 instruction-tuned models. They propose a comprehensive evaluation framework integrating precision, rank recovery, and directional error metrics, validated through human annotation, Top-1/3 accuracy, semantic panel aggregation, and zero-model baselines. Experimental results reveal average Acc@1 and Acc@3 scores of 0.683 and 0.892, respectively, with 50.9% of errors concentrated among neighboring values. The study further identifies eight recurrent, highly asymmetric directional confusions across models, highlighting potential biases that may distort higher-level value profiling.
0 citationsRead paper
REGARD: Regional Affective Differences in Large Language Models
This study addresses the limitations of existing approaches that rely solely on unidimensional sentiment polarity to assess large language models’ (LLMs) attitudes toward political and geopolitical entities, which fails to capture the nuanced complexity of regional affective expression. Introducing the valence–arousal–dominance (VAD) three-dimensional affect model to this domain for the first time, we systematically analyze affective disparities across 19 LLMs toward 500 post-Soviet targets. Leveraging large-scale model querying, dual-LLM automated scoring (GPT-4o-mini and Qwen3.6-35B-A3B), human annotation validation, and Ward hierarchical clustering, we identify three distinct behavioral clusters transcending model origin, architecture, and scale. Our analysis further reveals a strong negative correlation (r = –0.81) between generic response tendencies and arousal levels, demonstrating that the VAD framework effectively captures affective intensity and dominance dimensions overlooked by conventional methods, thereby enabling deeper structural insights into LLM emotional expression.
0 citationsRead paper
Portable models as a replacement for industrial heuristics in compiler optimizations
This work addresses the challenge that lightweight compilers and source-to-source tools struggle to reuse the sophisticated inlining heuristics of mature compilers like GCC or LLVM due to their reliance on complex intermediate representations and analysis infrastructures. To bridge this gap, the paper introduces the first portable inlining prediction framework that leverages diagnostic outputs from production compilers as supervision signals. By extracting call-site features through AST normalization and constructing a lightweight structured IR, the approach trains tabular models—such as CatBoost—that can be directly compiled into pure C code without runtime dependencies. Evaluated on a dataset of 330,000 call sites, the model achieves a ROC-AUC of 0.928 and PR-AUC of 0.713; with threshold tuning, it attains an F1 score of 0.729 while reducing the false positive rate to 0.084, thereby enabling the first practical transfer of industrial-grade inlining decisions to resource-constrained systems.
0 citationsRead paper
Cascade Classification of Dermoscopic Images of Skin Neoplasms with Controllable Sensitivity and External Clinical Validation
This study addresses the limited generalization and miscalibration of dermoscopic image classification models when transferring from international public datasets (e.g., ISIC Archive) to real-world clinical settings (e.g., Russian clinical data). To overcome the constraints of conventional single-stage argmax classification, the authors propose a tunable-sensitivity, two-stage cascaded framework: an initial benign–malignant binary screening followed by fine-grained subclassification of malignant cases, with adjustable decision thresholds aligned to clinical diagnostic logic. Evaluated consistently across ViT-B/16, Swin-S, ConvNeXt-S, and EfficientNetV2-S architectures, the approach achieves internal test ROC-AUC scores of 0.952–0.966 for binary classification. Although performance declines on external clinical data, the cascaded design significantly improves the macro F1 score for ViT-B/16 and effectively reduces missed diagnoses of malignant cases.
0 citationsRead paper
Recent publications
Latest Papers
CPL: A Compact C-like Systems Language with Explicit Low-Level Control
该论文介绍了一种紧凑的类C系统语言CPL，通过简化语法并结合新语言特性，旨在探索其在编译器实验中的便捷性，并展示了与主流C编译器相当的性能。
0 citationsRead paper
Which Values Do LLMs Confuse? A Schwartz-Based Recognition Study
This study addresses the limited reliability of large language models in identifying Schwartz’s ten basic human values, particularly their tendency to confuse adjacent value dimensions. The authors construct a balanced dataset comprising 1,000 Russian-language contextual prompts grounded in Schwartz’s theory and conduct the first systematic analysis of directional confusion patterns across 21 instruction-tuned models. They propose a comprehensive evaluation framework integrating precision, rank recovery, and directional error metrics, validated through human annotation, Top-1/3 accuracy, semantic panel aggregation, and zero-model baselines. Experimental results reveal average Acc@1 and Acc@3 scores of 0.683 and 0.892, respectively, with 50.9% of errors concentrated among neighboring values. The study further identifies eight recurrent, highly asymmetric directional confusions across models, highlighting potential biases that may distort higher-level value profiling.
0 citationsRead paper
REGARD: Regional Affective Differences in Large Language Models
This study addresses the limitations of existing approaches that rely solely on unidimensional sentiment polarity to assess large language models’ (LLMs) attitudes toward political and geopolitical entities, which fails to capture the nuanced complexity of regional affective expression. Introducing the valence–arousal–dominance (VAD) three-dimensional affect model to this domain for the first time, we systematically analyze affective disparities across 19 LLMs toward 500 post-Soviet targets. Leveraging large-scale model querying, dual-LLM automated scoring (GPT-4o-mini and Qwen3.6-35B-A3B), human annotation validation, and Ward hierarchical clustering, we identify three distinct behavioral clusters transcending model origin, architecture, and scale. Our analysis further reveals a strong negative correlation (r = –0.81) between generic response tendencies and arousal levels, demonstrating that the VAD framework effectively captures affective intensity and dominance dimensions overlooked by conventional methods, thereby enabling deeper structural insights into LLM emotional expression.
0 citationsRead paper
Portable models as a replacement for industrial heuristics in compiler optimizations
This work addresses the challenge that lightweight compilers and source-to-source tools struggle to reuse the sophisticated inlining heuristics of mature compilers like GCC or LLVM due to their reliance on complex intermediate representations and analysis infrastructures. To bridge this gap, the paper introduces the first portable inlining prediction framework that leverages diagnostic outputs from production compilers as supervision signals. By extracting call-site features through AST normalization and constructing a lightweight structured IR, the approach trains tabular models—such as CatBoost—that can be directly compiled into pure C code without runtime dependencies. Evaluated on a dataset of 330,000 call sites, the model achieves a ROC-AUC of 0.928 and PR-AUC of 0.713; with threshold tuning, it attains an F1 score of 0.729 while reducing the false positive rate to 0.084, thereby enabling the first practical transfer of industrial-grade inlining decisions to resource-constrained systems.
0 citationsRead paper
Cascade Classification of Dermoscopic Images of Skin Neoplasms with Controllable Sensitivity and External Clinical Validation
This study addresses the limited generalization and miscalibration of dermoscopic image classification models when transferring from international public datasets (e.g., ISIC Archive) to real-world clinical settings (e.g., Russian clinical data). To overcome the constraints of conventional single-stage argmax classification, the authors propose a tunable-sensitivity, two-stage cascaded framework: an initial benign–malignant binary screening followed by fine-grained subclassification of malignant cases, with adjustable decision thresholds aligned to clinical diagnostic logic. Evaluated consistently across ViT-B/16, Swin-S, ConvNeXt-S, and EfficientNetV2-S architectures, the approach achieves internal test ROC-AUC scores of 0.952–0.966 for binary classification. Although performance declines on external clinical data, the cascaded design significantly improves the macro F1 score for ViT-B/16 and effectively reduces missed diagnoses of malignant cases.
0 citationsRead paper