arXiv subjects

Seorin Kim

Publications and source records attributed to Seorin Kim.

4 recordsLinked to original sources

arXiv2026

One in Eight OpenAlex Abstracts Has Integrity Issues

Scientific abstracts are increasingly used as primary data in computational metascience research, yet the quality of these abstracts in widely used bibliographic databases has not been systematically examined. We assess the integrity of 10,000 randomly sampled English-language journal abstracts from OpenAlex using a two-stage annotation protocol combining human expert review and large language model classification. We identify seven distinct failure modes and find that 12\% of abstracts have integrity issues, with insufficient content and misplaced metadata being the most prevalent. We discuss implications for downstream research and describe a forthcoming community portal to support collective annotation efforts.

cs.DL↗

arXiv2026

Turning Citation Networks Inside Out: Studying Science Using Content-Based Knowledge Graphs from LLM-Derived Taxonomies

Scientific fields are often mapped using citations and metadata, despite knowledge being transmitted primarily through content. We introduce an 'inside-out' approach that reconstructs field structure directly from text by representing each paper as a small set of interpretable knowledge components. Using a large language model to induce domain-specific taxonomies and label papers, each publication is encoded as a triplet of measure, data type, and research-question type. These triplets define a knowledge graph with edges weighted by shared papers. Applied to 617 studies on intergenerational wealth mobility, the graph reveals a stable methodological backbone centered on regression-based mobility measures, alongside substantial temporal variation in component recombination. We further utilize normalized betweenness-to-connectivity ratios to identify components and pairings that act as structural bridges disproportionate to their prevalence. This content-derived, taxonomy-driven mapping complements citation-based approaches by exposing the evolving architecture of methods, data, and questions that define a field.

cs.SI↗

arXiv2025

KLAAD: Refining Attention Mechanisms to Reduce Societal Bias in Generative Language Models

Large language models (LLMs) often exhibit societal biases in their outputs, prompting ethical concerns regarding fairness and harm. In this work, we propose KLAAD (KL-Attention Alignment Debiasing), an attention-based debiasing framework that implicitly aligns attention distributions between stereotypical and anti-stereotypical sentence pairs without directly modifying model weights. KLAAD introduces a composite training objective combining Cross-Entropy, KL divergence, and Triplet losses, guiding the model to consistently attend across biased and unbiased contexts while preserving fluency and coherence. Experimental evaluation of KLAAD demonstrates improved bias mitigation on both the BBQ and BOLD benchmarks, with minimal impact on language modeling quality. The results indicate that attention-level alignment offers a principled solution for mitigating bias in generative language models.

cs.CL↗

arXiv2025

Early evidence of how LLMs outperform traditional systems on OCR/HTR tasks for historical records

We explore the ability of two LLMs -- GPT-4o and Claude Sonnet 3.5 -- to transcribe historical handwritten documents in a tabular format and compare their performance to traditional OCR/HTR systems: EasyOCR, Keras, Pytesseract, and TrOCR. Considering the tabular form of the data, two types of experiments are executed: one where the images are split line by line and the other where the entire scan is used as input. Based on CER and BLEU, we demonstrate that LLMs outperform the conventional OCR/HTR methods. Moreover, we also compare the evaluated CER and BLEU scores to human evaluations to better judge the outputs of whole-scan experiments and understand influential factors for CER and BLEU. Combining judgments from all the evaluation metrics, we conclude that two-shot GPT-4o for line-by-line images and two-shot Claude Sonnet 3.5 for whole-scan images yield the transcriptions of the historical records most similar to the ground truth.

cs.CV↗