Distinguishing AI-generated code from human-written code is becoming crucial for tasks such as authorship attribution, content tracking, and misuse detection. Based on this, N-gram-based watermarking schemes have emerged as prominent, which inject secret watermarks to be detected during the generation. However, their robustness in code content remains insufficiently evaluated. Most claims rely solely on defenses against simple code transformations or code optimizations as a simulation of attack, creating a questionable sense of robustness. In contrast, more sophisticated schemes already exist in the software engineering world, e.g., code obfuscation, which significantly alters code while preserving functionality. Although obfuscation is commonly used to protect intellectual property or evade software scanners, the robustness of code watermarking techniques against such transformations remains largely unexplored. In this work, we formally model the code obfuscation and prove the impossibility of N-gram-based watermarking's robustness with only one intuitive and experimentally verified assumption, distribution consistency, satisfied. Given the original false positive rate of the watermarking detection, the ratio that the detector failed on the watermarked code after obfuscation will increase to 1 - fpr. The experiments have been performed on three SOTA watermarking schemes, two LLMs, two programming languages, four code benchmarks, and four obfuscators. Among them, all watermarking detectors show coin-flipping detection abilities on obfuscated codes (AUROC tightly surrounds 0.5). Among all models, watermarking schemes, and datasets, both programming languages own obfuscators that can achieve attack effects with no detection AUROC higher than 0.6 after the attack. Based on the theoretical and practical observations, we also proposed a potential path of robust code watermarking.
Detecting AI-generated code, deepfakes, and other synthetic content is an emerging research challenge. As code generated by Large Language Models (LLMs) becomes more common, identifying the specific model behind each sample is increasingly important. This paper presents the first systematic study of LLM authorship attribution for C programs. We released CodeT5-Authorship, a novel model that uses only the encoder layers from the original CodeT5 encoder-decoder architecture, discarding the decoder to focus on classification. Our model's encoder output (first token) is passed through a two-layer classification head with GELU activation and dropout, producing a probability distribution over possible authors. To evaluate our approach, we introduce LLM-AuthorBench, a benchmark of 32,000 compilable C programs generated by eight state-of-the-art LLMs across diverse tasks. We compare our model to seven traditional ML classifiers and eight fine-tuned transformer models, including BERT, RoBERTa, CodeBERT, ModernBERT, DistilBERT, DeBERTa-V3, Longformer, and LoRA-fine-tuned Qwen2-1.5B. In binary classification, our model achieves 97.56% accuracy in distinguishing C programs generated by closely related models such as GPT-4.1 and GPT-4o, and 95.40% accuracy for multi-class attribution among five leading LLMs (Gemini 2.5 Flash, Claude 3.5 Haiku, GPT-4.1, Llama 3.3, and DeepSeek-V3). To support open science, we release the CodeT5-Authorship architecture, the LLM-AuthorBench benchmark, and all relevant Google Colab scripts on GitHub: https://github.com/LLMauthorbench/.
This paper presents the first comprehensive systematic review of literature on style-based composer identification and authorship attribution in symbolic music scores. Addressing the critical need for improved reliability and reproducibility in this field, the review rigorously analyzes 58 peer-reviewed papers published across various historical periods, with the search adapted to evolving terminology. The analysis critically assesses prevailing repertoires, computational approaches, and evaluation methodologies, highlighting significant challenges. It reveals that a substantial portion of existing research suffers from inadequate validation protocols and an over-reliance on simple accuracy metrics for often imbalanced datasets, which can undermine the credibility of attribution claims. The crucial role of robust metrics like Balanced Accuracy and rigorous cross-validation in ensuring trustworthy results is emphasized. The survey also details diverse feature representations and the evolution of machine learning models employed. Notable real-world authorship attribution cases, such as those involving works attributed to Bach, Josquin Desprez, and Lennon-McCartney, are specifically discussed, illustrating the opportunities and pitfalls of applying computational techniques to resolve disputed musical provenance. Based on these insights, a set of actionable guidelines for future research are proposed. These recommendations are designed to significantly enhance the reliability, reproducibility, and musicological validity of composer identification and authorship attribution studies, fostering more robust and interpretable computational stylistic analysis.
Alignment algorithms are widely used to align large language models (LLMs) to human users based on preference annotations. Typically these (often divergent) preferences are aggregated over a diverse set of users, resulting in fine-tuned models that are aligned to the ``average-user'' preference. Nevertheless, current models are used by individual users in very specific contexts and situations, emphasizing the need for user-dependent preference control. In this work we address the problem of personalizing LLM outputs to their users. We aim to generate customized responses tailored to specific individuals instead of generic outputs that emulate the collective voices of diverse populations. We propose HyPerAlign, an interpretable and sample-efficient hypothesis-driven personalization approach for LLM models. Given few-shot examples written by a particular user, we first infer hypotheses about their communication strategies, personality, and writing style, then prompt LLM models with these hypotheses and user-specific attributes to generate customized outputs. We conduct experiments on two different personalization tasks, namely authorship attribution and deliberative alignment, with datasets from diverse domains (news articles, blog posts, emails, jailbreaking benchmarks). Results demonstrate the superiority of hypothesis-driven LLM personalization compared to preference-based fine-tuning methods. For authorship attribution, HyPerAlign generations have consistently high win-rates (commonly $> 90\%$) against state-of-the-art preference fine-tuning approaches across diverse user profiles and LLM models. For deliberative alignment, the helpfulness of LLM models is improved by up to $70\%$ on average. Overall, HyPerAlign represents an interpretable and sample-efficient strategy for the personalization of LLM models to individual users.
Open Large Language Models (OLLMs) are increasingly leveraged in generative AI applications, posing new challenges for detecting their outputs. We propose OpenTuringBench, a new benchmark based on OLLMs, designed to train and evaluate machine-generated text detectors on the Turing Test and Authorship Attribution problems. OpenTuringBench focuses on a representative set of OLLMs, and features a number of challenging evaluation tasks, including human/machine-manipulated texts, out-of-domain texts, and texts from previously unseen models. We also provide OTBDetector, a contrastive learning framework to detect and attribute OLLM-based machine-generated texts. Results highlight the relevance and varying degrees of difficulty of the OpenTuringBench tasks, with our detector achieving remarkable capabilities across the various tasks and outperforming most existing detectors. Resources are available on the OpenTuringBench Hugging Face repository at https://huggingface.co/datasets/MLNTeam-Unical/OpenTuringBench
Traditionally, authorship attribution (AA) tasks relied on statistical data analysis and classification based on stylistic features extracted from texts. In recent years, pre-trained language models (PLMs) have attracted significant attention in text classification tasks. However, although they demonstrate excellent performance on large-scale short-text datasets, their effectiveness remains under-explored for small samples, particularly in AA tasks. Additionally, a key challenge is how to effectively leverage PLMs in conjunction with traditional feature-based methods to advance AA research. In this study, we aimed to significantly improve performance using an integrated integrative ensemble of traditional feature-based and modern PLM-based methods on an AA task in a small sample. For the experiment, we used two corpora of literary works to classify 10 authors each. The results indicate that BERT is effective, even for small-sample AA tasks. Both BERT-based and classifier ensembles outperformed their respective stand-alone models, and the integrated ensemble approach further improved the scores significantly. For the corpus that was not included in the pre-training data, the integrated ensemble improved the F1 score by approximately 14 points, compared to the best-performing single model. Our methodology provides a viable solution for the efficient use of the ever-expanding array of data processing tools in the foreseeable future.
For a long time, the authorship of the Federalist Papers had been a subject of inquiry and debate, not only by linguists and historians but also by statisticians. In what was arguably the first Bayesian case study, Mosteller and Wallace (1963) provided the first statistical evidence for attributing all disputed papers to Madison. Our paper revisits this historical dataset but from a lens of modern language models, both small and large. We review some of the more popular Large Language Model (LLM) tools and examine them from a statistical point of view in the context of text classification. We investigate whether, without any attempt to fine-tune, the general embedding constructs can be useful for stylometry and attribution. We explain differences between various word/phrase embeddings and discuss how to aggregate them in a document. Contrary to our expectations, we exemplify that dimension expansion with word embeddings may not always be beneficial for attribution relative to dimension reduction with topic embeddings. Our experiments demonstrate that default LLM embeddings (even after manual fine-tuning) may not consistently improve authorship attribution accuracy. Instead, Bayesian analysis with topic embeddings trained on ``function words" yields superior out-of-sample classification performance. This suggests that traditional (small) statistical language models, with their interpretability and solid theoretical foundation, can offer significant advantages in authorship attribution tasks. The code used in this analysis is available at github.com/sowonjeong/slm-to-llm
We propose a new approach for the authorship attribution task that leverages the various linguistic representations learned at different layers of pre-trained transformer-based models. We evaluate our approach on three datasets, comparing it to a state-of-the-art baseline in in-domain and out-of-domain scenarios. We found that utilizing various transformer layers improves the robustness of authorship attribution models when tested on out-of-domain data, resulting in new state-of-the-art results. Our analysis gives further insights into how our model's different layers get specialized in representing certain stylistic features that benefit the model when tested out of the domain.
The recent development and use of generative AI (GenAI) has signaled a significant shift in research activities such as brainstorming, proposal writing, dissemination, and even reviewing. This has raised questions about how to balance the seemingly productive uses of GenAI with ethical concerns such as authorship and copyright issues, use of biased training data, lack of transparency, and impact on user privacy. To address these concerns, many Higher Education Institutions (HEIs) have released institutional guidance for researchers. To better understand the guidance that is being provided we report findings from a thematic analysis of guidelines from thirty HEIs in the United States that are classified as R1 or 'very high research activity.' We found that guidance provided to researchers: (1) asks them to refer to external sources of information such as funding agencies and publishers to keep updated and use institutional resources for training and education; (2) asks them to understand and learn about specific GenAI attributes that shape research such as predictive modeling, knowledge cutoff date, data provenance, and model limitations, and educate themselves about ethical concerns such as authorship, attribution, privacy, and intellectual property issues; and (3) includes instructions on how to acknowledge sources and disclose the use of GenAI, how to communicate effectively about their GenAI use, and alerts researchers to long term implications such as over reliance on GenAI, legal consequences, and risks to their institutions from GenAI use. Overall, guidance places the onus of compliance on individual researchers making them accountable for any lapses, thereby increasing their responsibility.
Large Language Models (LLMs) are currently used extensively to generate code by professionals and students, motivating the development of tools to detect LLM-generated code for applications such as academic integrity and cybersecurity. We address this authorship attribution problem as a binary classification task along with feature identification and extraction. We propose new Discretized Nested Bigram Frequency features on source code groups of various sizes. Compared to prior work, improvements are obtained by representing sparse information in dense membership bins. Experimental evaluation demonstrated that our approach significantly outperformed a commonly used GPT code-detection API and baseline features, with accuracy exceeding 96% compared to 72% and 79% respectively in detecting GPT-rewritten Java code fragments for 976 files with GPT 3.5 and GPT4 using 12 features. We also outperformed three prior works on code author identification in a 40-author dataset. Our approach scales well to larger data sets, and we achieved 99% accuracy and 0.999 AUC for 76,089 files and over 1,000 authors with GPT 4o using 227 features.
Masked language modeling has become a widely adopted unsupervised technique to pre-train language models. However, the process of selecting tokens for masking is random, and the percentage of masked tokens is typically fixed for the entire training process. In this paper, we propose to adjust the masking ratio and to decide which tokens to mask based on a novel task-informed anti-curriculum learning scheme. First, we harness task-specific knowledge about useful and harmful tokens in order to determine which tokens to mask. Second, we propose a cyclic decaying masking ratio, which corresponds to an anti-curriculum schedule (from hard to easy). We exemplify our novel task-informed anti-curriculum by masking (TIACBM) approach across three diverse downstream tasks: sentiment analysis, text classification by topic, and authorship attribution. Our findings suggest that TIACBM enhances the ability of the model to focus on key task-relevant features, contributing to statistically significant performance gains across tasks. We release our code at https://github.com/JarcaAndrei/TIACBM.
Natural Language Processing (NLP) for lesser-resourced languages faces persistent challenges, including limited datasets, inherited biases from high-resource languages, and the need for domain-specific solutions. This study addresses these gaps for Modern Greek through three key contributions. First, we evaluate the performance of open-source (Llama-70b) and closed-source (GPT-4o mini) large language models (LLMs) on seven core NLP tasks with dataset availability, revealing task-specific strengths, weaknesses, and parity in their performance. Second, we expand the scope of Greek NLP by reframing Authorship Attribution as a tool to assess potential data usage by LLMs in pre-training, with high 0-shot accuracy suggesting ethical implications for data provenance. Third, we showcase a legal NLP case study, where a Summarize, Translate, and Embed (STE) methodology outperforms the traditional TF-IDF approach for clustering \emph{long} legal texts. Together, these contributions provide a roadmap to advance NLP in lesser-resourced languages, bridging gaps in model evaluation, task innovation, and real-world impact.
With the advent of Web 2.0, the development in social technology coupled with global communication systematically brought positive and negative impacts to society. Copyright claims and Author identification are deemed crucial as there has been a considerable amount of increase in content violation owing to the lack of proper ethics in society. The Author's attribution in both English and Romanized Sinhala became a major requirement in the last few decades. As an area largely unexplored, particularly within the context of Romanized Sinhala, the research contributes significantly to the field of computational linguistics. The proposed author attribution system offers a unique approach, allowing for the comparison of only two sets of text: suspect author and anonymous text, a departure from traditional methodologies which often rely on larger corpora. This work focuses on using the numerical representation of various pairs of the same and different authors allowing for, the model to train on these representations as opposed to text, this allows for it to apply to a multitude of authors and contexts, given that the suspected author text, and the anonymous text are of reasonable quality. By expanding the scope of authorship attribution to encompass diverse linguistic contexts, the work contributes to fostering trust and accountability in digital communication, especially in Sri Lanka. This research presents a pioneering approach to author attribution in both English and Romanized Sinhala, addressing a critical need for content verification and intellectual property rights enforcement in the digital age.
Source code authorship attribution is important in software forensics, plagiarism detection, and protecting software patch integrity. Existing techniques often rely on supervised machine learning, which struggles with generalization across different programming languages and coding styles due to the need for large labeled datasets. Inspired by recent advances in natural language authorship analysis using large language models (LLMs), which have shown exceptional performance without task-specific tuning, this paper explores the use of LLMs for source code authorship attribution. We present a comprehensive study demonstrating that state-of-the-art LLMs can successfully attribute source code authorship across different languages. LLMs can determine whether two code snippets are written by the same author with zero-shot prompting, achieving a Matthews Correlation Coefficient (MCC) of 0.78, and can attribute code authorship from a small set of reference code snippets via few-shot learning, achieving MCC of 0.77. Additionally, LLMs show some adversarial robustness against misattribution attacks. Despite these capabilities, we found that naive prompting of LLMs does not scale well with a large number of authors due to input token limitations. To address this, we propose a tournament-style approach for large-scale attribution. Evaluating this approach on datasets of C++ (500 authors, 26,355 samples) and Java (686 authors, 55,267 samples) code from GitHub, we achieve classification accuracy of up to 65% for C++ and 68.7% for Java using only one reference per author. These results open new possibilities for applying LLMs to code authorship attribution in cybersecurity and software engineering.
Human trafficking (HT) remains a critical issue, with traffickers increasingly leveraging online escort advertisements (ads) to advertise victims anonymously. Existing detection methods, including Authorship Attribution (AA), often center on text-based analyses and neglect the multimodal nature of online escort ads, which typically pair text with images. To address this gap, we introduce MATCHED, a multimodal dataset of 27,619 unique text descriptions and 55,115 unique images collected from the Backpage escort platform across seven U.S. cities in four geographical regions. Our study extensively benchmarks text-only, vision-only, and multimodal baselines for vendor identification and verification tasks, employing multitask (joint) training objectives that achieve superior classification and retrieval performance on in-distribution and out-of-distribution (OOD) datasets. Integrating multimodal features further enhances this performance, capturing complementary patterns across text and images. While text remains the dominant modality, visual data adds stylistic cues that enrich model performance. Moreover, text-image alignment strategies like CLIP and BLIP2 struggle due to low semantic overlap and vague connections between the modalities of escort ads, with end-to-end multimodal training proving more robust. Our findings emphasize the potential of multimodal AA (MAA) to combat HT, providing LEAs with robust tools to link ads and disrupt trafficking networks.
In an era where cyberattacks increasingly target the software supply chain, the ability to accurately attribute code authorship in binary files is critical to improving cybersecurity measures. We propose OCEAN, a contrastive learning-based system for function-level authorship attribution. OCEAN is the first framework to explore code authorship attribution on compiled binaries in an open-world and extreme scenario, where two code samples from unknown authors are compared to determine if they are developed by the same author. To evaluate OCEAN, we introduce new realistic datasets: CONAN, to improve the performance of authorship attribution systems in real-world use cases, and SNOOPY, to increase the robustness of the evaluation of such systems. We use CONAN to train our model and evaluate on SNOOPY, a fully unseen dataset, resulting in an AUROC score of 0.86 even when using high compiler optimizations. We further show that CONAN improves performance by 7% compared to the previously used Google Code Jam dataset. Additionally, OCEAN outperforms previous methods in their settings, achieving a 10% improvement over state-of-the-art SCS-Gan in scenarios analyzing source code. Furthermore, OCEAN can detect code injections from an unknown author in a software update, underscoring its value for securing software supply chains.
The increasing prevalence of AI-generated content alongside human-written text underscores the need for reliable discrimination methods. To address this challenge, we propose a novel framework with textual embeddings from Pre-trained Language Models (PLMs) to distinguish AI-generated and human-authored text. Our approach utilizes Embedding Fusion to integrate semantic information from multiple Language Models, harnessing their complementary strengths to enhance performance. Through extensive evaluation across publicly available diverse datasets, our proposed approach demonstrates strong performance, achieving classification accuracy greater than 96% and a Matthews Correlation Coefficient (MCC) greater than 0.93. This evaluation is conducted on a balanced dataset of texts generated from five well-known Large Language Models (LLMs), highlighting the effectiveness and robustness of our novel methodology.
Authorship attribution aims to identify the origin or author of a document. Traditional approaches have heavily relied on manual features and fail to capture long-range correlations, limiting their effectiveness. Recent advancements leverage text embeddings from pre-trained language models, which require significant fine-tuning on labeled data, posing challenges in data dependency and limited interpretability. Large Language Models (LLMs), with their deep reasoning capabilities and ability to maintain long-range textual associations, offer a promising alternative. This study explores the potential of pre-trained LLMs in one-shot authorship attribution, specifically utilizing Bayesian approaches and probability outputs of LLMs. Our methodology calculates the probability that a text entails previous writings of an author, reflecting a more nuanced understanding of authorship. By utilizing only pre-trained models such as Llama-3-70B, our results on the IMDb and blog datasets show an impressive 85\% accuracy in one-shot authorship classification across ten authors. Our findings set new baselines for one-shot authorship analysis using LLMs and expand the application scope of these models in forensic linguistics. This work also includes extensive ablation studies to validate our approach.
This paper evaluates the performance of Large Language Models (LLMs) in authorship attribution and authorship verification tasks for Latin texts of the Patristic Era. The study showcases that LLMs can be robust in zero-shot authorship verification even on short texts without sophisticated feature engineering. Yet, the models can also be easily "mislead" by semantics. The experiments also demonstrate that steering the model's authorship analysis and decision-making is challenging, unlike what is reported in the studies dealing with high-resource modern languages. Although LLMs prove to be able to beat, under certain circumstances, the traditional baselines, obtaining a nuanced and truly explainable decision requires at best a lot of experimentation.
The rapid expansion of medical informatics literature presents significant challenges in synthesizing and analyzing research trends. This study introduces a novel dataset derived from the Medical Informatics Europe (MIE) Conference proceedings, addressing the need for sophisticated analytical tools in the field. Utilizing the Triple-A software, we extracted and processed metadata and abstract from 4,606 articles published in the "Studies in Health Technology and Informatics" journal series, focusing on MIE conferences from 1996 onwards. Our methodology incorporated advanced techniques such as affiliation parsing using the TextRank algorithm. The resulting dataset, available in JSON format, offers a comprehensive view of bibliometric details, extracted topics, and standardized affiliation information. Analysis of this data revealed interesting patterns in Digital Object Identifier usage, citation trends, and authorship attribution across the years. Notably, we observed inconsistencies in author data and a brief period of linguistic diversity in publications. This dataset represents a significant contribution to the medical informatics community, enabling longitudinal studies of research trends, collaboration network analyses, and in-depth bibliometric investigations. By providing this enriched, structured resource spanning nearly three decades of conference proceedings, we aim to facilitate novel insights and advancements in the rapidly evolving field of medical informatics.