Recent advances in Large Language Models (LLMs) have shown promise in function-level code generation, yet repository-level software engineering tasks remain challenging. Current solutions predominantly rely on proprietary LLM agents, which introduce unpredictability and limit accessibility, raising concerns about data privacy and model customization. This paper investigates whether open-source LLMs can effectively address repository-level tasks without requiring agent-based approaches. We demonstrate this is possible by enabling LLMs to comprehend functions and files within codebases through their semantic information and structural dependencies. To this end, we introduce Code Graph Models (CGMs), which integrate repository code graph structures into the LLM's attention mechanism and map node attributes to the LLM's input space using a specialized adapter. When combined with an agentless graph RAG framework, our approach achieves a 43.00% resolution rate on the SWE-bench Lite benchmark using the open-source Qwen2.5-72B model. This performance ranks first among open weight models, second among methods with open-source systems, and eighth overall, surpassing the previous best open-source model-based method by 12.33%.

With the emergence of ChatGPT, Transformer models have significantly advanced text classification and related tasks. Decoder-only models such as Llama exhibit strong performance and flexibility, yet they suffer from inefficiency on inference due to token-by-token generation, and their effectiveness in text classification tasks heavily depends on prompt quality. Moreover, their substantial GPU resource requirements often limit widespread adoption. Thus, the question of whether smaller language models are capable of effectively handling text classification tasks emerges as a topic of significant interest. However, the selection of appropriate models and methodologies remains largely underexplored. In this paper, we conduct a comprehensive evaluation of prompt engineering and supervised fine-tuning methods for transformer-based text classification. Specifically, we focus on practical industrial scenarios, including email classification, legal document categorization, and the classification of extremely long academic texts. We examine the strengths and limitations of smaller models, with particular attention to both their performance and their efficiency in Video Random-Access Memory (VRAM) utilization, thereby providing valuable insights for the local deployment and application of compact models in industrial settings.

While understanding the knowledge boundaries of LLMs is crucial to prevent hallucination, research on the knowledge boundaries of LLMs has predominantly focused on English. In this work, we present the first study to analyze how LLMs recognize knowledge boundaries across different languages by probing their internal representations when processing known and unknown questions in multiple languages. Our empirical studies reveal three key findings: 1) LLMs' perceptions of knowledge boundaries are encoded in the middle to middle-upper layers across different languages. 2) Language differences in knowledge boundary perception follow a linear structure, which motivates our proposal of a training-free alignment method that effectively transfers knowledge boundary perception ability across languages, thereby helping reduce hallucination risk in low-resource languages; 3) Fine-tuning on bilingual question pair translation further enhances LLMs' recognition of knowledge boundaries across languages. Given the absence of standard testbeds for cross-lingual knowledge boundary analysis, we construct a multilingual evaluation suite comprising three representative types of knowledge boundary data. Our code and datasets are publicly available at https://github.com/DAMO-NLP-SG/LLM-Multilingual-Knowledge-Boundaries.

A United States federal judge has ruled that the company Anthropic made "fair use" of the books it utilised to train artificial intelligence (AI) tools without the permission of the authors. The favourable ruling comes at a time when the impacts of AI are being discussed by regulators and policymakers, and the industry is using its political influence to push for a loose regulatory framework. "Like any reader aspiring to be a writer, Anthropic's LLMs [large language models] trained upon works not to race ahead and replicate or supplant them -- but to turn a hard corner and create something different," US District Judge William Alsup said. A group of authors had filed a class-action lawsuit alleging that Anthropic's use of their work to train its chatbot, Claude, without their consent was illegal. He accepted Anthropic's claim that the AI's output was "exceedingly transformative" and therefore fell under the "fair use" protections.

"The training use was a fair use," senior district judge William Alsup wrote in a summary judgement order released late Monday evening. "The technology at issue was among the most transformative many of us will see in our lifetimes," Alsup wrote. "Judge Alsup found that training an LLM is transformative use--even when there is significant memorization. He specifically rejected the argument that what humans do when reading and memorizing is different in kind from what computers do when training an LLM." Anthropic is the first artificial intelligence company to win this kind of battle, but the victory comes with a large asterisk attached. While Alsup found that Anthropic's training was fair use, he ruled that the authors could take Anthropic to trial over pirating their works.

In the era of large language models (LLMs), code benchmarks have become an important research area in software engineering and are widely used by practitioners. These benchmarks evaluate the performance of LLMs on specific code-related tasks, such as code understanding and generation. A critical step in constructing code benchmarks is the design of prompts. However, as existing code benchmarks typically rely on a single prompt template per task, they are prone to the issue of prompt sensitivity, where minor prompt variations could result in substantial performance variations, leading to unreliable evaluations of model capabilities. While previous studies have explored prompt sensitivity, their experimental designs and findings are limited to traditional natural language processing (NLP) tasks. In this paper, we present an empirical study to investigate prompt sensitivity in code benchmarks. We first propose a general framework that modifies prompt templates in a manner that preserves both their semantics and their structure as much as possible. Based on the framework, we conduct extensive experiments across eight code benchmark tasks on 10 representative open-source LLMs, with each task featuring 100 semantically similar prompt templates. We then analyze the evaluation results using various statistical metrics, focusing on both absolute and relative model performance. Our findings suggest that even slight prompt variations can lead to significant shifts in performance. Additionally, we observe that such variations can introduce inconsistencies in the performance rankings across different models. These insights highlight the need for considering prompt sensitivity when designing future code benchmarks, to ensure more reliable and accurate evaluation of LLM capabilities.

Despite progress in controllable symbolic music generation, data scarcity remains a challenge for certain control modalities. Composer-style music generation is a prime example, as only a few pieces per composer are available, limiting the modeling of both styles and fundamental music elements (e.g., melody, chord, rhythm). In this paper, we investigate how general music knowledge learned from a broad corpus can enhance the mastery of specific composer styles, with a focus on piano piece generation. Our approach follows a two-stage training paradigm. First, we pre-train a REMI-based music generation model on a large corpus of pop, folk, and classical music. Then, we fine-tune it on a small, human-verified dataset from four renowned composers, namely Bach, Mozart, Beethoven, and Chopin, using a lightweight adapter module to condition the model on style indicators. To evaluate the effectiveness of our approach, we conduct both objective and subjective evaluations on style accuracy and musicality. Experimental results demonstrate that our method outperforms ablations and baselines, achieving more precise composer-style modeling and better musical aesthetics. Additionally, we provide observations on how the model builds music concepts from the generality pre-training and refines its stylistic understanding through the mastery fine-tuning.

Tutoring improves student achievement, but identifying and studying what tutoring actions are most associated with student learning at scale based on audio transcriptions is an open research problem. This present study investigates the feasibility and scalability of using generative AI to identify and evaluate specific tutor moves in real-life math tutoring. We analyze 50 randomly selected transcripts of college-student remote tutors assisting middle school students in mathematics. Using GPT-4, GPT-4o, GPT-4-turbo, Gemini-1.5-pro, and LearnLM, we assess tutors' application of two tutor skills: delivering effective praise and responding to student math errors. All models reliably detected relevant situations, for example, tutors providing praise to students (94-98% accuracy) and a student making a math error (82-88% accuracy) and effectively evaluated the tutors' adherence to tutoring best practices, aligning closely with human judgments (83-89% and 73-77%, respectively). We propose a cost-effective prompting strategy and discuss practical implications for using large language models to support scalable assessment in authentic settings. This work further contributes LLM prompts to support reproducibility and research in AI-supported learning.

This draft is a working document, having a summary of nighty-four (94) papers with additional sections on Traceability of Software Requirements (Section 4), Formal Methods and Its Tools (Section 5), Unifying Theories of Programming (UTP) and Theory of Institutions (Section 6). Please refer to abstract of [7,8]. Key difference of this draft from our recently anticipated ones with similar titles, i.e. AACS 2025 [7] and SAIV 2025 [8] is: [7] is a two page submission to ADAPT Annual Conference, Ireland. Submitted on 18th of March, 2025, it went through the light-weight blind review and accepted for poster presentation. Conference was held on 15th of May, 2025; [8] is a nine page paper with additional nine pages of references and summary tables, submitted to Symposium on AI Verification (SAIV 2025) on 24th of April, 2025. It went through rigorous review process. The uploaded version on arXiv.org [8] is the improved one of the submission, after addressing the specific suggestions to improve the paper.

Part-of-speech (POS) tagging remains a foundational component in natural language processing pipelines, particularly critical for historical text analysis at the intersection of computational linguistics and digital humanities. Despite significant advancements in modern large language models (LLMs) for ancient languages, their application to Medieval Romance languages presents distinctive challenges stemming from diachronic linguistic evolution, spelling variations, and labeled data scarcity. This study systematically investigates the central determinants of POS tagging performance across diverse corpora of Medieval Occitan, Medieval Spanish, and Medieval French texts, spanning biblical, hagiographical, medical, and dietary domains. Through rigorous experimentation, we evaluate how fine-tuning approaches, prompt engineering, model architectures, decoding strategies, and cross-lingual transfer learning techniques affect tagging accuracy. Our results reveal both notable limitations in LLMs' ability to process historical language variations and non-standardized spelling, as well as promising specialized techniques that effectively address the unique challenges presented by low-resource historical languages.