This paper describes our system for SemEval-2026 Task 13 on detecting machine-generated code. We fine-tune small encoder-only models for detecting human-written versus machine-generated code and for identifying which large language model (LLM) family was used to obtain code. We find that a strong, general-purpose model (ModernBERT) outperforms models specifically pre-trained for the code domain. In the official evaluation, our system ranks 5th on subtask B and 6th on subtask C. Our detailed analysis reveals that comments and other natural language text that is part of the code snippets provide valuable information for identifying the LLM family that generated it. Moreover, we show that the embeddings of our finetuned ModernBERT do not distinguish well between LLM families, but they cluster human-written code by programming language.

This paper describes our system for the SciVQA 2025 Shared Task on Scientific Visual Question Answering. Our system employs an ensemble of two Multimodal Large Language Models and various few-shot example retrieval strategies. The model and few-shot setting are selected based on the figure and question type. We also select answers based on the models’ confidence levels. On the blind test data, our system ranks third out of seven with an average F1 score of 85.12 across ROUGE-1, ROUGE-L, and BERTS. Our code is publicly available.

The scientific literature is growing rapidly, making it hard to keep track of the state-of-the-art. Systematic literature reviews (SLRs) aim to identify and evaluate all relevant papers on a topic. After retrieving a set of candidate papers, the abstract screening phase determines initial relevance. To date, abstract screening methods using large language models (LLMs) focus on binary classification settings; existing question answering (QA) based ranking approaches suffer from error propagation. LLMs offer a unique opportunity to evaluate the SLR’s inclusion and exclusion criteria, yet, existing benchmarks do not provide them exhaustively. We manually extract these criteria as well as research questions for 57 SLRs, mostly in the medical domain, enabling principled comparisons between approaches. Moreover, we propose LGAR, a zero-shot LLM Guided Abstract Ranker composed of an LLM based graded relevance scorer and a dense re-ranker. Our extensive experiments show that LGAR outperforms existing QA-based methods by 5-10 pp. in mean average precision. Our code and data is publicly available.