small language model

ACADEMIC PERMISSIBILITY IN LLM-ASSISTED EFL WRITING: HOW CONTEXTUAL JUSTIFICATIONS SHAPE STUDENTS' MORAL JUDGMENTS IN A WITHIN-SUBJECTS VIGNETTE SURVEY

AbstractThis paper investigates EFL students’ perceptions of ethical acceptability judgments of using Large Language Models (LLMs) into academic writing. In contrast to the simplistic view of acceptable/unacceptable use of LLMs, the present study models how specific contextual justifications shape students’ moral evaluations of LLM-assisted writing. A cross-sectional within-subject design was used with 220 third-year EFL students at three public universities in Laghouat, Algeria, to rate the ethical acceptability of LLMs use to complete four writing assignments. Under a neutral baseline condition, participants assessed the use of LLM for completing the four assignments, as well as four single condition contexts (disclosure, accuracy verification, syllabus permission, and learning intent). Their ratings were analyzed using Δ-effect scores (conditional minus baseline) to quantify condition effects above baseline and to describe task-level differences in ethical acceptability. Contextual conditions increased ethical acceptability to different extents, with learning intent producing the largest positive shift (ΔM ≈ 1.6, d ≈ 0.7) and disclosure exerting only a small, non-robust effect (ΔM ≈ 0.2, d ≈ 0.1). The baseline ratings also followed a clear gradient, which saw AI-assisted proofreading as the most acceptable while paraphrasing was consistently least acceptable. Theoretically, the study adds value by supporting a conditional-ethics perspective because it demonstrates how students’ moral considerations of LLMs use are dependent on learning-oriented and epistemically responsible frameworks rather than procedural cues like disclosure or syllabus permissions with no behavioral consequences. Practically, the paper advices that AI policies and pedagogy should be designed to encourage learning intent and verification activities as opposed to disclosure as a standalone requirement.Keywords: Academic integrity; large language models (LLMs); EFL writing; conditional ethics

ACADEMIC PERMISSIBILITY IN LLM-ASSISTED EFL WRITING: HOW CONTEXTUAL JUSTIFICATIONS SHAPE STUDENTS' MORAL JUDGMENTS IN A WITHIN-SUBJECTS VIGNETTE SURVEY

AbstractThis paper investigates EFL students’ perceptions of ethical acceptability judgments of using Large Language Models (LLMs) into academic writing. In contrast to the simplistic view of acceptable/unacceptable use of LLMs, the present study models how specific contextual justifications shape students’ moral evaluations of LLM-assisted writing. A cross-sectional within-subject design was used with 220 third-year EFL students at three public universities in Laghouat, Algeria, to rate the ethical acceptability of LLMs use to complete four writing assignments. Under a neutral baseline condition, participants assessed the use of LLM for completing the four assignments, as well as four single condition contexts (disclosure, accuracy verification, syllabus permission, and learning intent). Their ratings were analyzed using Δ-effect scores (conditional minus baseline) to quantify condition effects above baseline and to describe task-level differences in ethical acceptability. Contextual conditions increased ethical acceptability to different extents, with learning intent producing the largest positive shift (ΔM ≈ 1.6, d ≈ 0.7) and disclosure exerting only a small, non-robust effect (ΔM ≈ 0.2, d ≈ 0.1). The baseline ratings also followed a clear gradient, which saw AI-assisted proofreading as the most acceptable while paraphrasing was consistently least acceptable. Theoretically, the study adds value by supporting a conditional-ethics perspective because it demonstrates how students’ moral considerations of LLMs use are dependent on learning-oriented and epistemically responsible frameworks rather than procedural cues like disclosure or syllabus permissions with no behavioral consequences. Practically, the paper advices that AI policies and pedagogy should be designed to encourage learning intent and verification activities as opposed to disclosure as a standalone requirement.Keywords: Academic integrity; large language models (LLMs); EFL writing; conditional ethics

Can Large Language Models Serve as Data Analysts? A Multi-Agent Assisted Approach for Qualitative Data Analysis

Context: Manual qualitative data analysis is time-intensive and can compromise validity and replicability, affecting analysis design, implementation, and reporting. Large Language Models (LLMs) enable human-bot collaboration in Software Engineering (SE), but their potential for qualitative data analysis in SE remains largely unexplored. Objective: The objective of this study is to design and develop an LLM-based multi-agent system that synergizes human decision support with AI to automate various qualitative data analysis approaches. Methods: We used LLM-based multi-agents systems to assist the qualitative data analysis process, deploying 27 agents, each responsible for a specific task, such as text summarization, initial code generation, and extracting themes and patterns. Results: The main findings are: (1) the LLM-based multi-agent system accelerates the qualitative data analysis process, (2) the system effectively automates tasks such as text summarization, initial code generation, and theme extraction, and (3) the publicly accessible code facilitates validation and further evaluation. Conclusion: The proposed LLM-based multi-agent system automates qualitative data analysis process, creating opportunities for researchers and practitioners. Future improvements focus on enhancing multilingual performance and integrating continuous expert feedback. The source code of proposed system and system details can be found here: https://github.com/GPT-Laboratory/Qualitative-Analysis-with-an-LLM-Based-Agentts

LLM-Based Multi-Agent Systems for Code Generation: A Multi-Vocal Literature Review

Large Language Models (LLMs) have enabled multi-agent systems to perform autonomous code generation for complex tasks. Despite the recent growth in research and industrial applications in this area, there is little work on synthesizing evidence from both academic and industrial sources to capture the current state of research on LLM-based multi-agent systems for code generation. To this end, we conducted a Multi-Vocal Literature Review (MLR), combining insights from both academia and industry, including peer-reviewed studies and grey literature. The aim of this study is to systematically synthesize and analyze existing knowledge on LLM-based multi-agent systems for code generation. Specifically, the review examines the motivations for their use, employed benchmarks and models, key challenges, proposed solutions, and potential directions for future research. We selected and reviewed 114 studies, and the key findings are: 1) the identified reasons for adopting multi-agent systems for code generation were classified into nine categories; 2) the models and evaluation benchmarks utilized across the studies were systematically analyzed to provide a structured overview of commonly adopted LLM configurations and assessment practices; 3) the reported challenges and corresponding solutions were synthesized into six main categories and 26 subcategories; and 4) future research directions were identified and organized into six main categories and 18 subcategories. The results of this MLR will assist researchers and practitioners in pursuing further studies and supporting the real-world adoption of multi-agent systems in industrial settings.

Autonomous Agents in Software Development: A Vision Paper

Large Language Models (LLM) and Generative Pre-trained Transformers (GPT), are reshaping the field of Software Engineering (SE). They enable innovative methods for executing many software engineering tasks, including automated code generation, debugging, maintenance, etc. However, only a limited number of existing works have thoroughly explored the potential of GPT agents in SE. This vision paper inquires about the role of GPT-based agents in SE. Our vision is to leverage the capabilities of multiple GPT agents to contribute to SE tasks and to propose an initial road map for future work. We argue that multiple GPT agents can perform creative and demanding tasks far beyond coding and debugging. GPT agents can also do project planning, requirements engineering, and software design. These can be done through high-level descriptions given by the human developer. We have shown in our initial experimental analysis for simple software (e.g., Snake Game, Tic-Tac-Toe, Notepad) that multiple GPT agents can produce high-quality code and document it carefully. We argue that it shows a promise of unforeseen efficiency and will dramatically reduce lead-times. To this end, we intend to expand our efforts to understand how we can scale these autonomous capabilities further.

Reproductive phasiRNAs: At the core of small RNA-mediated male fertility in monocots

HSRM: Hidden-State Reward Models for Test-Time Verification

Large language models can often generate plausible mathematical reasoning traces, but reliably identifying the correct solution among multiple candidates remains a key challenge. Existing test-time reasoning pipelines typically rely on text-based verifiers that re-read each generated solution, making verification an expensive component of inference. Prior work has shown, however, that LLMs often encode correctness-related signals in their internal representations, including awareness of when their own answers are likely to be wrong. Building on this observation, we introduce HSRM, a lightweight hidden-state reward model that verifies candidate solutions by directly reading the generator's internal representations rather than re-processing its text. HSRM extracts hidden states from a frozen generator at reasoning-step boundaries and uses a small Transformer encoder to rank candidates. It is trained from self-generated trajectories with outcome labels, requiring neither human-written process supervision nor a large pretrained verifier. Across four mathematical reasoning benchmarks, HSRM matches or outperforms a 55M-parameter text-only energy verifier in 15 of 16 generator--dataset settings while using only about 2M parameters, providing an efficient alternative to text-only verification by reusing representations already computed during generation.

A.X K2 Technical Report

We introduce A.X K2, a 688B-parameter Mixture-of-Experts (MoE) language model trained from scratch as a high-performance foundation for \emph{agentic} applications. Trained on approximately 8.5T tokens---fewer than its predecessor, A.X K1---on a smaller but higher-quality mixture with substantially expanded agentic and software-engineering data, it nonetheless improves over A.X K1 across the board, by over 30 percentage points on some benchmarks, reflecting large gains in token efficiency. To support long contexts efficiently, we introduce Sparse Gated Attention (SGA), which combines sparse attention with gated attention, and adopt Gated Norm (GN) to stabilize large-scale training. SGA is trained natively at 128K through a \emph{sparse} indexer warmup that optimizes the indexer against its own sparse top-$k$ selection rather than the dense attention distribution, making adaptation markedly cheaper: each query reads only 2,048 positions, yet long-context quality is unchanged and A.X K2 scores 94.6 on RULER out to 256K. The outlier suppression of GN in turn keeps 4-bit NVFP4 serving within one point of FP8 accuracy. A simple yet effective Think-Fusion recipe further lets users switch between thinking and non-thinking modes within a single unified model. Extensive evaluations show that A.X K2 performs competitively against strong open-weight baselines, matching or exceeding them on math and Korean-language benchmarks.

ECGQuest: Benchmarking and Fine-Tuning Language Models for Electrocardiography

Electrocardiogram (ECG) interpretation requires knowledge of cardiology, electrophysiology, clinical diagnosis, ECG waveforms, signal acquisition, and instrumentation. Existing language-model benchmarks, however, primarily assess broad medical knowledge or interpretation of individual ECG signals and images rather than the broader contextual knowledge required for ECG interpretation. We developed ECGQuest, a literature-grounded resource for evaluating and fine-tuning ECG-specific language models. A GPT-4o-based pipeline generated questions from 23 ECG references and Computing in Cardiology proceedings from 2003-2025. The final dataset contains 10,904 unique True/False questions paired with their negated forms (21,808 Q&A pairs). We evaluated three commercial and 20 open-source language models on a held-out test set in a zero-shot setting. Five open-source models with 7-14B parameters were fine-tuned using Low-Rank Adaptation, with BERT and BiomedBERT included as supervised encoder baselines. Generalization was assessed on ECG-related subsets of MedMCQA and MedQA converted to binary True/False questions using official answer keys. Zero-shot accuracy on ECGQuest ranged from 49.5% to 74.4%, with GPT-5 performing best. General-purpose models outperformed medically specialized models, several models showed strong True/False bias, and encoder baselines performed near chance. Fine-tuning improved all open-source models by 6.5-14.1%. Fine-tuned DeepSeek-R1-Distill-Qwen-14B reached 76.3% accuracy, while a five-model voting ensemble reached 78.5%. On MedMCQA and MedQA, fine-tuning mainly benefited weaker or class-biased models and did not consistently improve strong base models. ECGQuest provides a reproducible benchmark for contextual ECG knowledge and shows that parameter-efficient fine-tuning can make smaller language models competitive with substantially larger commercial models.

Error-Type-Aware Loss Reweighting for Robust Named Entity Recognition with Noisy LLM Labels

Large language models are increasingly used to annotate datasets for training smaller, task-specialized models such as named entity recognition. While this method yields effective models, it assumes that the synthetic dataset is correctly annotated. In this work, we find that (i) current fine-tuning processes simply ignore LLM-introduced annotation noise, resulting in degraded performance and (ii) existing noise-robust losses are not transferable to sequence labeling because annotation noise in named entity recognition is heterogeneous: for example, missing mentions and type errors affect the training signal in different ways. Treating all noisy tokens equally in noise-robust losses and applying a single reweighing criterion for all may therefore remove useful supervision or reinforce incorrect labels. To address this limitation, we propose error-type-aware loss reweighting for NER, which introduces separate reweighing rules for different types of potentially erroneous tokens. Our approach is simple and efficient, does not require additional training resources, and improves F1 by 0.8 - 2.0 percentage points on dataset-level average for noise levels between 15% and 40%, with a maximum improvement of 4.6 percentage points with 24.1% noise on Wikigold.

Calibrating Small Language Models for Claim Check-Worthiness Detection

Assessing claim check-worthiness is an essential first step in automated fact-checking pipelines. This work is motivated by a real deployment challenge at an early-stage startup: running large language models (LLMs) over every incoming claim is cost- and latency-prohibitive, yet smaller models sacrifice accuracy. We propose NN-PPI, a pointwise extension of Prediction-Powered Inference (PPI) that calibrates model predictions at inference time as a lightweight post-hoc layer, without re-training the underlying model. NN-PPI achieves weighted F1 gains ranging from 12% to 33.80% depending on the size and performance of the baseline model, bringing SLMs on par with larger LLMs. Beyond few-shot SLMs, NN-PPI further improves a production-deployed fine-tuned model, demonstrating that residual calibration is complementary to supervised fine-tuning. By recovering LLM-level accuracy from models that are an order of magnitude cheaper to serve, it makes accurate check-worthiness detection substantially cheaper to operate at scale. Our code and data can be found at https://anonymous.4open.science/r/arr-claim-worthiness-F237.

LCoT-GV: Graph Attention Networks for Verifying Long Reasoning Chains in Large Language Models

Large Reasoning Models produce Long Chains-of-Thought (LCoTs) which involve breaking down the problem into smaller reasoning steps before reaching the conclusion. However, these steps often contain contradictions, unsupported inferences, or irrelevant steps, even when the final answer is correct. We propose Long Chain-of-Thought Graph Verifier (LCoT-GV), a graph-based framework that represents LCoTs as reasoning graphs. Each node in the graph represents a reasoning step and the edges encode semantic and logical relations. A Graph Attention Network is then trained to predict chain-of-thought correctness from the reasoning graph. We construct a new graph-oriented verification dataset from multiple reasoning benchmarks for question answering in various domains. The results show that our method is competitive with the most similar approaches.