Tag
The author fact-checks their previous post on NeurIPS 2026 position track decisions and Pangram AI detection, correcting earlier errors and providing updated benchmarks and reports.
ICLR 2027 faces a de-anonymization incident where submissions were exposed to program committee members, raising privacy and security concerns for academic conferences.
A local AI pipeline named Academic-AI-Literature-Reviewer-Ollama generates fully cited APA 7 literature reviews with a verification layer that prevents citation fabrication by matching quotes against source PDFs.
OpenAI faces controversy for allegedly using academic researchers' progress to solve the Millennium Prize Navier-Stokes problem first, with the achievement involving massive compute and raising questions about academic ethics.
The paper theoretically analyzes how repeated paraphrasing can align AI-generated text with human writing distributions, deriving convergence rates and implications for reliable AI-text detection.
This paper presents a benchmark for classifying research papers into good (highly cited) and non-good (retracted) using textual features from titles and abstracts, with models like SVM and neural networks achieving up to 91.12% accuracy.
Ghent University suspended Nathan Cofnas after he reported plagiarism in the Arday scandal, with rector Petra De Sutter making the decision to potentially fire him.
This paper evaluates commercial AI detectors in academic settings, finding high false-positive rates on AI-assisted human writing and near-total evasion via humanizers, concluding detector scores should not be standalone misconduct evidence.
Denmark introduces mandatory oral defenses for high school students' written assignments to counter AI cheating, along with recommendations for screen monitoring and firewalls. Education groups welcome the measures but call for long-term solutions as AI tools evolve.
Discusses the emerging practice of using LLMs to remix existing papers and evade plagiarism detection, warning of a collapse in academic ethics.
This paper proposes a zero-shot detection pipeline to characterize human-likeness in AI-generated poetry, aiming to identify attributes that distinguish human from machine poems and improve detection methods.
A NeurIPS reviewer reports encountering a paper and rebuttals that appear entirely LLM-generated, expressing frustration and seeking advice on how to evaluate such submissions.
This paper investigates how well LLMs can detect their own generated content in educational contexts, finding that detection accuracy varies by task type and is unreliable for short-answer questions.
A Brown University professor highlights a chart showing a 'wake-up call' about AI cheating in academic assignments, raising concerns about policing writing tasks with tools like GPT4All and the need for school-provided LLM gateways.
A Brown University professor observed that most students likely cheated on a take-home exam using AI, as evidenced by score differences when the final was held in-person. The story highlights ongoing debates about AI and academic integrity.
Brown University economics professor Roberto Serrano, suspecting AI cheating after record-high scores on a take-home midterm, administered an in-person final exam; scores plummeted by 50%, revealing widespread AI-assisted cheating among Ivy League students.
Professor Roberto Serrano at Brown University detected at least 50 students cheating using AI on a midterm exam, sparking a debate about academic integrity in higher education.
Medical students are reportedly using a popular research tool to produce misleading studies, raising concerns about academic integrity and research quality.
A student expresses frustration that their entirely human-written paper was flagged as AI-generated by plagiarism checkers, highlighting the flaws of current AI detection tools in academic settings.
A French physicist and media personality had their doctorate revoked following an investigation into plagiarism.