{"topic":"reasoning-models","sort":"top","limit":25,"offset":0,"count":25,"papers":[{"id":"arxiv:2501.12948","kind":"article","title":"DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning","authors":["Daya Guo","Dejian Yang","Haowei Zhang","Junxiao Song","Peiyi Wang","Qihao Zhu","Runxin Xu","Ruoyu Zhang","Shirong Ma","Xiao Bi","Xiaokang Zhang","Xingkai Yu","Yu Wu","Z. F. Wu","Zhibin Gou","Zhihong Shao","Zhuoshu Li","Ziyi Gao","Aixin Liu","Bing Xue","Bingxuan Wang","Bowen Wu","Bei Feng","Chengda Lu","Chenggang Zhao"],"author_count":100,"year":2025,"published":"2025-09-17","venue":"Nature","cited_by":1687,"doi":"10.1038/s41586-025-09422-z","arxiv_id":"2501.12948","nct_id":null,"source_url":"https://doi.org/10.1038/s41586-025-09422-z","archive_url":"https://space-kitty.com/papers/arxiv:2501.12948","updated":"2026-01-04","pmid":"40962978","pmcid":null,"openalex_id":"W4414281281","pdf_url":"https://www.nature.com/articles/s41586-025-09422-z.pdf","abstract":"General reasoning represents a long-standing and formidable challenge in artificial intelligence. Recent breakthroughs, exemplified by large language models (LLMs) and chain-of-thought prompting, have achieved considerable success on foundational reasoning tasks. However, this success is heavily contingent upon extensive human-annotated demonstrations, and models' capabilities are still insufficient for more complex problems. Here we show that the reasoning abilities of LLMs can be incentivized through pure reinforcement learning (RL), obviating the need for human-labeled reasoning trajectories. The proposed RL framework facilitates the emergent development of advanced reasoning patterns, such as self-reflection, verification, and dynamic strategy adaptation. Consequently, the trained model achieves superior performance on verifiable tasks such as mathematics, coding competitions, and STEM fields, surpassing its counterparts trained via conventional supervised learning on human demonstrations. Moreover, the emergent reasoning patterns exhibited by these large-scale models can be systematically harnessed to guide and enhance the reasoning capabilities of smaller models.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":"cc-by","is_open_access":true,"peer_reviewed":null,"categories":["Topic Modeling"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2201.11903","kind":"preprint","title":"Chain-of-Thought Prompting Elicits Reasoning in Large Language Models","authors":["Wei, Jason","Wang, Xuezhi","Dale Schuurmans","Maarten Bosma","Brian Ichter","Fei Xia","Ed H.","Quoc Viet Le","Denny Zhou"],"author_count":9,"year":2022,"published":"2022-01-28","venue":null,"cited_by":4209,"doi":null,"arxiv_id":"2201.11903","nct_id":null,"source_url":"http://arxiv.org/abs/2201.11903","archive_url":"https://space-kitty.com/papers/arxiv:2201.11903","updated":"2023-01-10","pmid":null,"pmcid":null,"openalex_id":"W4221143046","pdf_url":"https://arxiv.org/pdf/2201.11903","abstract":"We explore how generating a chain of thought -- a series of intermediate reasoning steps -- significantly improves the ability of large language models to perform complex reasoning. In particular, we show how such reasoning abilities emerge naturally in sufficiently large language models via a simple method called chain of thought prompting, where a few chain of thought demonstrations are provided as exemplars in prompting. Experiments on three large language models show that chain of thought prompting improves performance on a range of arithmetic, commonsense, and symbolic reasoning tasks. The empirical gains can be striking. For instance, prompting a 540B-parameter language model with just eight chain of thought exemplars achieves state of the art accuracy on the GSM8K benchmark of math word problems, surpassing even finetuned GPT-3 with a verifier.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":true,"peer_reviewed":false,"categories":["Topic Modeling"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2203.11171","kind":"preprint","title":"Self-Consistency Improves Chain of Thought Reasoning in Language Models","authors":["Xuezhi Wang","Wei, Jason","Dale Schuurmans","Quoc Viet Le","Ed H.","Narang, Sharan","Aakanksha Chowdhery","Zhou, Denny"],"author_count":8,"year":2022,"published":"2022-03-21","venue":null,"cited_by":809,"doi":null,"arxiv_id":"2203.11171","nct_id":null,"source_url":"http://arxiv.org/abs/2203.11171","archive_url":"https://space-kitty.com/papers/arxiv:2203.11171","updated":"2023-03-07","pmid":null,"pmcid":null,"openalex_id":"W4221161695","pdf_url":"https://arxiv.org/pdf/2203.11171","abstract":"Chain-of-thought prompting combined with pre-trained large language models has achieved encouraging results on complex reasoning tasks. In this paper, we propose a new decoding strategy, self-consistency, to replace the naive greedy decoding used in chain-of-thought prompting. It first samples a diverse set of reasoning paths instead of only taking the greedy one, and then selects the most consistent answer by marginalizing out the sampled reasoning paths. Self-consistency leverages the intuition that a complex reasoning problem typically admits multiple different ways of thinking leading to its unique correct answer. Our extensive empirical evaluation shows that self-consistency boosts the performance of chain-of-thought prompting with a striking margin on a range of popular arithmetic and commonsense reasoning benchmarks, including GSM8K (+17.9%), SVAMP (+11.0%), AQuA (+12.2%), StrategyQA (+6.4%) and ARC-challenge (+3.9%).","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":true,"peer_reviewed":false,"categories":["Topic Modeling"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2210.03629","kind":"preprint","title":"ReAct: Synergizing Reasoning and Acting in Language Models","authors":["Yao, Shunyu","Jeffrey W. Zhao","Dian Yu","Nan Du","Izhak Shafran J","Karthik Narasimhan","Yuan Cao"],"author_count":7,"year":2022,"published":"2022-10-06","venue":null,"cited_by":607,"doi":null,"arxiv_id":"2210.03629","nct_id":null,"source_url":"http://arxiv.org/abs/2210.03629","archive_url":"https://space-kitty.com/papers/arxiv:2210.03629","updated":"2023-03-10","pmid":null,"pmcid":null,"openalex_id":"W4304195432","pdf_url":"https://arxiv.org/pdf/2210.03629","abstract":"While large language models (LLMs) have demonstrated impressive capabilities across tasks in language understanding and interactive decision making, their abilities for reasoning (e.g. chain-of-thought prompting) and acting (e.g. action plan generation) have primarily been studied as separate topics. In this paper, we explore the use of LLMs to generate both reasoning traces and task-specific actions in an interleaved manner, allowing for greater synergy between the two: reasoning traces help the model induce, track, and update action plans as well as handle exceptions, while actions allow it to interface with external sources, such as knowledge bases or environments, to gather additional information. We apply our approach, named ReAct, to a diverse set of language and decision making tasks and demonstrate its effectiveness over state-of-the-art baselines, as well as improved human interpretability and trustworthiness over methods without reasoning or acting components. Concretely, on question answering (HotpotQA) and fact verification (Fever), ReAct overcomes issues of hallucination and error propagation prevalent in chain-of-thought reasoning by interacting with a simple Wikipedia API, and generates human-like task-solving trajectories that are more interpretable than baselines without reasoning traces. On two interactive decision making benchmarks (ALFWorld and WebShop), ReAct outperforms imitation and reinforcement learning methods by an absolute success rate of 34% and 10% respectively, while being prompted with only one or two in-context examples. Project site with code: https://react-lm.github.io","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":true,"peer_reviewed":false,"categories":["Ethics and Social Impacts of AI"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2207.08143","kind":"article","title":"Can large language models reason about medical questions?","authors":["Valentin Liévin","Christoffer Hother","Andreas Geert Motzfeldt","Ole Winther"],"author_count":4,"year":2024,"published":"2024-03-01","venue":"Patterns","cited_by":343,"doi":"10.1016/j.patter.2024.100943","arxiv_id":"2207.08143","nct_id":null,"source_url":"https://doi.org/10.1016/j.patter.2024.100943","archive_url":"https://space-kitty.com/papers/arxiv:2207.08143","updated":"2023-12-24","pmid":"38487804","pmcid":null,"openalex_id":"W4392359953","pdf_url":"http://www.cell.com/article/S2666389924000424/pdf","abstract":"Although large language models (LLMs) often produce impressive outputs, it remains unclear how they perform in real-world scenarios requiring strong reasoning skills and expert domain knowledge. We set out to investigate whether close- and open-source models (GPT-3.5, LLama-2, etc.) can be applied to answer and reason about difficult real-world-based questions. We focus on three popular medical benchmarks (MedQA-USMLE, MedMCQA, and PubMedQA) and multiple prompting scenarios: Chain-of-Thought (CoT, think step-by-step), few-shot and retrieval augmentation. Based on an expert annotation of the generated CoTs, we found that InstructGPT can often read, reason and recall expert knowledge. Last, by leveraging advances in prompt engineering (few-shot and ensemble methods), we demonstrated that GPT-3.5 not only yields calibrated predictive distributions, but also reaches the passing score on three datasets: MedQA-USMLE 60.2%, MedMCQA 62.7% and PubMedQA 78.2%. Open-source models are closing the gap: Llama-2 70B also passed the MedQA-USMLE with 62.5% accuracy.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":"cc-by","is_open_access":true,"peer_reviewed":null,"categories":["Artificial Intelligence in Healthcare and Education"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2205.10625","kind":"preprint","title":"Least-to-Most Prompting Enables Complex Reasoning in Large Language Models","authors":["Denny Zhou","Nathanael Schärli A","Le Hou","Wei, Jason","Scales, Nathan","Xuezhi Wang","Dale Schuurmans","Claire Murong Cui","Olivier Bousquet","Quoc Viet Le","Ed H."],"author_count":11,"year":2022,"published":"2022-05-21","venue":null,"cited_by":379,"doi":null,"arxiv_id":"2205.10625","nct_id":null,"source_url":"http://arxiv.org/abs/2205.10625","archive_url":"https://space-kitty.com/papers/arxiv:2205.10625","updated":"2023-04-16","pmid":null,"pmcid":null,"openalex_id":"W4281483047","pdf_url":"https://arxiv.org/pdf/2205.10625","abstract":"Chain-of-thought prompting has demonstrated remarkable performance on various natural language reasoning tasks. However, it tends to perform poorly on tasks which requires solving problems harder than the exemplars shown in the prompts. To overcome this challenge of easy-to-hard generalization, we propose a novel prompting strategy, least-to-most prompting. The key idea in this strategy is to break down a complex problem into a series of simpler subproblems and then solve them in sequence. Solving each subproblem is facilitated by the answers to previously solved subproblems. Our experimental results on tasks related to symbolic manipulation, compositional generalization, and math reasoning reveal that least-to-most prompting is capable of generalizing to more difficult problems than those seen in the prompts. A notable finding is that when the GPT-3 code-davinci-002 model is used with least-to-most prompting, it can solve the compositional generalization benchmark SCAN in any split (including length split) with an accuracy of at least 99% using just 14 exemplars, compared to only 16% accuracy with chain-of-thought prompting. This is particularly noteworthy because neural-symbolic models in the literature that specialize in solving SCAN are trained on the entire training set containing over 15,000 examples. We have included prompts for all the tasks in the Appendix.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":true,"peer_reviewed":false,"categories":["Topic Modeling"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"doi:10.3390/healthcare11202776","kind":"review","title":"Leveraging Generative AI and Large Language Models: A Comprehensive Roadmap for Healthcare Integration","authors":["Ping Yu","Hua Xu","Xia Hu","Chao Deng"],"author_count":4,"year":2023,"published":"2023-10-20","venue":"Healthcare","cited_by":261,"doi":"10.3390/healthcare11202776","arxiv_id":null,"nct_id":null,"source_url":"https://doi.org/10.3390/healthcare11202776","archive_url":"https://space-kitty.com/papers/doi:10.3390/healthcare11202776","updated":null,"pmid":"37893850","pmcid":null,"openalex_id":"W4387809804","pdf_url":"https://www.mdpi.com/2227-9032/11/20/2776/pdf?version=1697783532","abstract":"Generative artificial intelligence (AI) and large language models (LLMs), exemplified by ChatGPT, are promising for revolutionizing data and information management in healthcare and medicine. However, there is scant literature guiding their integration for non-AI professionals. This study conducts a scoping literature review to address the critical need for guidance on integrating generative AI and LLMs into healthcare and medical practices. It elucidates the distinct mechanisms underpinning these technologies, such as Reinforcement Learning from Human Feedback (RLFH), including few-shot learning and chain-of-thought reasoning, which differentiates them from traditional, rule-based AI systems. It requires an inclusive, collaborative co-design process that engages all pertinent stakeholders, including clinicians and consumers, to achieve these benefits. Although global research is examining both opportunities and challenges, including ethical and legal dimensions, LLMs offer promising advancements in healthcare by enhancing data management, information retrieval, and decision-making processes. Continued innovation in data acquisition, model fine-tuning, prompt strategy development, evaluation, and system implementation is imperative for realizing the full potential of these technologies. Organizations should proactively engage with these technologies to improve healthcare quality, safety, and efficiency, adhering to ethical and legal guidelines for responsible application.","abstract_license":"CC-BY (publisher license, via OpenAlex)","license":"cc-by","is_open_access":true,"peer_reviewed":null,"categories":["Artificial Intelligence in Healthcare and Education"],"details":null,"provenance":"openalex","attribution":"Abstract of \"Leveraging Generative AI and Large Language Models: A Comprehensive Roadmap for Healthcare Integration\" by Ping Yu, Hua Xu, Xia Hu et al., https://doi.org/10.3390/healthcare11202776, used under its Creative Commons Attribution license (CC-BY); markup removed, otherwise unmodified."},{"id":"arxiv:2312.03748","kind":"article","title":"Applying large language models and chain-of-thought for automatic scoring","authors":["Gyeong-Geon Lee","Ehsan Latif","Xuansheng Wu","Ninghao Liu","Xiaoming Zhaı"],"author_count":5,"year":2024,"published":"2024-02-27","venue":"Computers and Education Artificial Intelligence","cited_by":186,"doi":"10.1016/j.caeai.2024.100213","arxiv_id":"2312.03748","nct_id":null,"source_url":"https://doi.org/10.1016/j.caeai.2024.100213","archive_url":"https://space-kitty.com/papers/arxiv:2312.03748","updated":"2024-02-16","pmid":null,"pmcid":null,"openalex_id":"W4392202731","pdf_url":"https://arxiv.org/pdf/2312.03748","abstract":"This study investigates the application of large language models (LLMs), specifically GPT-3.5 and GPT-4, with Chain-of-Though (CoT) in the automatic scoring of student-written responses to science assessments. We focused on overcoming the challenges of accessibility, technical complexity, and lack of explainability that have previously limited the use of artificial intelligence-based automatic scoring tools among researchers and educators. With a testing dataset comprising six assessment tasks (three binomial and three trinomial) with 1,650 student responses, we employed six prompt engineering strategies to automatically score student responses. The six strategies combined zero-shot or few-shot learning with CoT, either alone or alongside item stem and scoring rubrics. Results indicated that few-shot (acc = .67) outperformed zero-shot learning (acc = .60), with 12.6% increase. CoT, when used without item stem and scoring rubrics, did not significantly affect scoring accuracy (acc = .60). However, CoT prompting paired with contextual item stems and rubrics proved to be a significant contributor to scoring accuracy (13.44% increase for zero-shot; 3.7% increase for few-shot). We found a more balanced accuracy across different proficiency categories when CoT was used with a scoring rubric, highlighting the importance of domain-specific reasoning in enhancing the effectiveness of LLMs in scoring tasks. We also found that GPT-4 demonstrated superior performance over GPT -3.5 in various scoring tasks when combined with the single-call greedy sampling or ensemble voting nucleus sampling strategy, showing 8.64% difference. Particularly, the single-call greedy sampling strategy with GPT-4 outperformed other approaches.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":"cc-by-nc-nd","is_open_access":true,"peer_reviewed":null,"categories":["Artificial Intelligence in Education"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2210.03493","kind":"preprint","title":"Automatic Chain of Thought Prompting in Large Language Models","authors":["Zhuosheng Zhang","Aston Zhang","Mu Li","Alex J. Smola"],"author_count":4,"year":2022,"published":"2022-10-07","venue":null,"cited_by":274,"doi":null,"arxiv_id":"2210.03493","nct_id":null,"source_url":"http://arxiv.org/abs/2210.03493","archive_url":"https://space-kitty.com/papers/arxiv:2210.03493","updated":"2022-10-07","pmid":null,"pmcid":null,"openalex_id":"W4304194220","pdf_url":"https://arxiv.org/pdf/2210.03493","abstract":"Large language models (LLMs) can perform complex reasoning by generating intermediate reasoning steps. Providing these steps for prompting demonstrations is called chain-of-thought (CoT) prompting. CoT prompting has two major paradigms. One leverages a simple prompt like \"Let's think step by step\" to facilitate step-by-step thinking before answering a question. The other uses a few manual demonstrations one by one, each composed of a question and a reasoning chain that leads to an answer. The superior performance of the second paradigm hinges on the hand-crafting of task-specific demonstrations one by one. We show that such manual efforts may be eliminated by leveraging LLMs with the \"Let's think step by step\" prompt to generate reasoning chains for demonstrations one by one, i.e., let's think not just step by step, but also one by one. However, these generated chains often come with mistakes. To mitigate the effect of such mistakes, we find that diversity matters for automatically constructing demonstrations. We propose an automatic CoT prompting method: Auto-CoT. It samples questions with diversity and generates reasoning chains to construct demonstrations. On ten public benchmark reasoning tasks with GPT-3, Auto-CoT consistently matches or exceeds the performance of the CoT paradigm that requires manual designs of demonstrations. Code is available at https://github.com/amazon-research/auto-cot","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":true,"peer_reviewed":false,"categories":["Topic Modeling"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"doi:10.3390/medicina60030445","kind":"review","title":"Integrating Retrieval-Augmented Generation with Large Language Models in Nephrology: Advancing Practical Applications","authors":["Jing Miao","Charat Thongprayoon","Supawadee Suppadungsuk","Oscar A. Garcia Valencia","Wisit Cheungpasitporn"],"author_count":5,"year":2024,"published":"2024-03-08","venue":"Medicina","cited_by":168,"doi":"10.3390/medicina60030445","arxiv_id":null,"nct_id":null,"source_url":"https://doi.org/10.3390/medicina60030445","archive_url":"https://space-kitty.com/papers/doi:10.3390/medicina60030445","updated":null,"pmid":"38541171","pmcid":null,"openalex_id":"W4392597393","pdf_url":"https://www.mdpi.com/1648-9144/60/3/445/pdf?version=1709877206","abstract":"The integration of large language models (LLMs) into healthcare, particularly in nephrology, represents a significant advancement in applying advanced technology to patient care, medical research, and education. These advanced models have progressed from simple text processors to tools capable of deep language understanding, offering innovative ways to handle health-related data, thus improving medical practice efficiency and effectiveness. A significant challenge in medical applications of LLMs is their imperfect accuracy and/or tendency to produce hallucinations-outputs that are factually incorrect or irrelevant. This issue is particularly critical in healthcare, where precision is essential, as inaccuracies can undermine the reliability of these models in crucial decision-making processes. To overcome these challenges, various strategies have been developed. One such strategy is prompt engineering, like the chain-of-thought approach, which directs LLMs towards more accurate responses by breaking down the problem into intermediate steps or reasoning sequences. Another one is the retrieval-augmented generation (RAG) strategy, which helps address hallucinations by integrating external data, enhancing output accuracy and relevance. Hence, RAG is favored for tasks requiring up-to-date, comprehensive information, such as in clinical decision making or educational applications. In this article, we showcase the creation of a specialized ChatGPT model integrated with a RAG system, tailored to align with the KDIGO 2023 guidelines for chronic kidney disease. This example demonstrates its potential in providing specialized, accurate medical advice, marking a step towards more reliable and efficient nephrology practices.","abstract_license":"CC-BY (publisher license, via OpenAlex)","license":"cc-by","is_open_access":true,"peer_reviewed":null,"categories":["Artificial Intelligence in Healthcare and Education"],"details":null,"provenance":"openalex","attribution":"Abstract of \"Integrating Retrieval-Augmented Generation with Large Language Models in Nephrology: Advancing Practical Applications\" by Jing Miao, Charat Thongprayoon, Supawadee Suppadungsuk et al., https://doi.org/10.3390/medicina60030445, used under its Creative Commons Attribution license (CC-BY); markup removed, otherwise unmodified."},{"id":"doi:10.1145/3690635","kind":"article","title":"Structured Chain-of-Thought Prompting for Code Generation","authors":["Jia Li","Ge Li","Yongmin Li","Zhi Gang Jin"],"author_count":4,"year":2024,"published":"2024-08-29","venue":"ACM Transactions on Software Engineering and Methodology","cited_by":150,"doi":"10.1145/3690635","arxiv_id":null,"nct_id":null,"source_url":"https://doi.org/10.1145/3690635","archive_url":"https://space-kitty.com/papers/doi:10.1145/3690635","updated":null,"pmid":null,"pmcid":null,"openalex_id":"W4401996408","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3690635","abstract":null,"abstract_license":null,"license":null,"is_open_access":true,"peer_reviewed":null,"categories":["Software Engineering Research"],"details":null,"provenance":"openalex","attribution":"Metadata and citation counts: OpenAlex (CC0)."},{"id":"arxiv:2310.08034","kind":"article","title":"Receive, Reason, and React: Drive as You Say, With Large Language Models in Autonomous Vehicles","authors":["Can Cui","Yunsheng Ma","Xu Cao","Wenqian Ye","Ziran Wang"],"author_count":5,"year":2024,"published":"2024-04-09","venue":"IEEE Intelligent Transportation Systems Magazine","cited_by":116,"doi":"10.1109/mits.2024.3381793","arxiv_id":"2310.08034","nct_id":null,"source_url":"https://doi.org/10.1109/mits.2024.3381793","archive_url":"https://space-kitty.com/papers/arxiv:2310.08034","updated":"2023-10-12","pmid":null,"pmcid":null,"openalex_id":"W4394595621","pdf_url":"https://arxiv.org/pdf/2310.08034","abstract":"The fusion of human-centric design and artificial intelligence (AI) capabilities has opened up new possibilities for next-generation autonomous vehicles that go beyond transportation. These vehicles can dynamically interact with passengers and adapt to their preferences. This paper proposes a novel framework that leverages Large Language Models (LLMs) to enhance the decision-making process in autonomous vehicles. By utilizing LLMs' linguistic and contextual understanding abilities with specialized tools, we aim to integrate the language and reasoning capabilities of LLMs into autonomous vehicles. Our research includes experiments in HighwayEnv, a collection of environments for autonomous driving and tactical decision-making tasks, to explore LLMs' interpretation, interaction, and reasoning in various scenarios. We also examine real-time personalization, demonstrating how LLMs can influence driving behaviors based on verbal commands. Our empirical results highlight the substantial advantages of utilizing chain-of-thought prompting, leading to improved driving decisions, and showing the potential for LLMs to enhance personalized driving experiences through ongoing verbal feedback. The proposed framework aims to transform autonomous vehicle operations, offering personalized support, transparent decision-making, and continuous learning to enhance safety and effectiveness. We achieve user-centric, transparent, and adaptive autonomous driving ecosystems supported by the integration of LLMs into autonomous vehicles.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":"cc-by-nc-sa","is_open_access":true,"peer_reviewed":null,"categories":["Autonomous Vehicle Technology and Safety"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2211.10435","kind":"preprint","title":"PAL: Program-aided Language Models","authors":["Luyu Gao","Aman Madaan","Shuyan Zhou","Uri Alon","Pengfei Liu","Yiming Yang","Jamie Callan","Graham Neubig"],"author_count":8,"year":2022,"published":"2022-11-18","venue":null,"cited_by":155,"doi":null,"arxiv_id":"2211.10435","nct_id":null,"source_url":"http://arxiv.org/abs/2211.10435","archive_url":"https://space-kitty.com/papers/arxiv:2211.10435","updated":"2023-01-27","pmid":null,"pmcid":null,"openalex_id":"W4309591663","pdf_url":"https://arxiv.org/pdf/2211.10435","abstract":"Large language models (LLMs) have recently demonstrated an impressive ability to perform arithmetic and symbolic reasoning tasks, when provided with a few examples at test time (\"few-shot prompting\"). Much of this success can be attributed to prompting methods such as \"chain-of-thought'', which employ LLMs for both understanding the problem description by decomposing it into steps, as well as solving each step of the problem. While LLMs seem to be adept at this sort of step-by-step decomposition, LLMs often make logical and arithmetic mistakes in the solution part, even when the problem is decomposed correctly. In this paper, we present Program-Aided Language models (PAL): a novel approach that uses the LLM to read natural language problems and generate programs as the intermediate reasoning steps, but offloads the solution step to a runtime such as a Python interpreter. With PAL, decomposing the natural language problem into runnable steps remains the only learning task for the LLM, while solving is delegated to the interpreter. We demonstrate this synergy between a neural LLM and a symbolic interpreter across 13 mathematical, symbolic, and algorithmic reasoning tasks from BIG-Bench Hard and other benchmarks. In all these natural language reasoning tasks, generating code using an LLM and reasoning using a Python interpreter leads to more accurate results than much larger models. For example, PAL using Codex achieves state-of-the-art few-shot accuracy on the GSM8K benchmark of math word problems, surpassing PaLM-540B which uses chain-of-thought by absolute 15% top-1. Our code and data are publicly available at http://reasonwithpal.com/ .","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":true,"peer_reviewed":false,"categories":["Topic Modeling"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"doi:10.3390/medicina60010148","kind":"review","title":"Chain of Thought Utilization in Large Language Models and Application in Nephrology","authors":["Jing Miao","Charat Thongprayoon","Supawadee Suppadungsuk","Pajaree Krisanapan","Yeshwanter Radhakrishnan","Wisit Cheungpasitporn"],"author_count":6,"year":2024,"published":"2024-01-13","venue":"Medicina","cited_by":102,"doi":"10.3390/medicina60010148","arxiv_id":null,"nct_id":null,"source_url":"https://doi.org/10.3390/medicina60010148","archive_url":"https://space-kitty.com/papers/doi:10.3390/medicina60010148","updated":null,"pmid":"38256408","pmcid":null,"openalex_id":"W4390880284","pdf_url":"https://www.mdpi.com/1648-9144/60/1/148/pdf?version=1705129367","abstract":"Chain-of-thought prompting enhances the abilities of large language models (LLMs) significantly. It not only makes these models more specific and context-aware but also impacts the wider field of artificial intelligence (AI). This approach broadens the usability of AI, increases its efficiency, and aligns it more closely with human thinking and decision-making processes. As we improve this method, it is set to become a key element in the future of AI, adding more purpose, precision, and ethical consideration to these technologies. In medicine, the chain-of-thought prompting is especially beneficial. Its capacity to handle complex information, its logical and sequential reasoning, and its suitability for ethically and context-sensitive situations make it an invaluable tool for healthcare professionals. Its role in enhancing medical care and research is expected to grow as we further develop and use this technique. Chain-of-thought prompting bridges the gap between AI's traditionally obscure decision-making process and the clear, accountable standards required in healthcare. It does this by emulating a reasoning style familiar to medical professionals, fitting well into their existing practices and ethical codes. While solving AI transparency is a complex challenge, the chain-of-thought approach is a significant step toward making AI more comprehensible and trustworthy in medicine. This review focuses on understanding the workings of LLMs, particularly how chain-of-thought prompting can be adapted for nephrology's unique requirements. It also aims to thoroughly examine the ethical aspects, clarity, and future possibilities, offering an in-depth view of the exciting convergence of these areas.","abstract_license":"CC-BY (publisher license, via OpenAlex)","license":"cc-by","is_open_access":true,"peer_reviewed":null,"categories":["Artificial Intelligence in Healthcare and Education"],"details":null,"provenance":"openalex","attribution":"Abstract of \"Chain of Thought Utilization in Large Language Models and Application in Nephrology\" by Jing Miao, Charat Thongprayoon, Supawadee Suppadungsuk et al., https://doi.org/10.3390/medicina60010148, used under its Creative Commons Attribution license (CC-BY); markup removed, otherwise unmodified."},{"id":"arxiv:2211.12588","kind":"preprint","title":"Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks","authors":["Wenhu Chen","Xueguang Ma","Xinyi Wang","William W. Cohen"],"author_count":4,"year":2022,"published":"2022-11-22","venue":null,"cited_by":125,"doi":null,"arxiv_id":"2211.12588","nct_id":null,"source_url":"http://arxiv.org/abs/2211.12588","archive_url":"https://space-kitty.com/papers/arxiv:2211.12588","updated":"2023-10-23","pmid":null,"pmcid":null,"openalex_id":"W4309953112","pdf_url":"https://arxiv.org/pdf/2211.12588","abstract":"Recently, there has been significant progress in teaching language models to perform step-by-step reasoning to solve complex numerical reasoning tasks. Chain-of-thoughts prompting (CoT) is by far the state-of-art method for these tasks. CoT uses language models to perform both reasoning and computation in the multi-step `thought' process. To disentangle computation from reasoning, we propose `Program of Thoughts' (PoT), which uses language models (mainly Codex) to express the reasoning process as a program. The computation is relegated to an external computer, which executes the generated programs to derive the answer. We evaluate PoT on five math word problem datasets (GSM, AQuA, SVAMP, TabMWP, MultiArith) and three financial-QA datasets (FinQA, ConvFinQA, TATQA) for both few-shot and zero-shot setups. Under both few-shot and zero-shot settings, PoT can show an average performance gain over CoT by around 12\\% across all the evaluated datasets. By combining PoT with self-consistency decoding, we can achieve SoTA performance on all math problem datasets and near-SoTA performance on financial datasets. All of our data and code are released in Github https://github.com/wenhuchen/Program-of-Thoughts","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":true,"peer_reviewed":false,"categories":["Topic Modeling"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2302.00923","kind":"preprint","title":"Multimodal Chain-of-Thought Reasoning in Language Models","authors":["Zhuosheng Zhang","Aston Zhang","Mu Li","Hai Yan Zhao","George Karypis","Alex J. Smola"],"author_count":6,"year":2023,"published":"2023-02-02","venue":null,"cited_by":110,"doi":null,"arxiv_id":"2302.00923","nct_id":null,"source_url":"http://arxiv.org/abs/2302.00923","archive_url":"https://space-kitty.com/papers/arxiv:2302.00923","updated":"2024-05-20","pmid":null,"pmcid":null,"openalex_id":"W4319165821","pdf_url":"https://arxiv.org/pdf/2302.00923","abstract":"Large language models (LLMs) have shown impressive performance on complex reasoning by leveraging chain-of-thought (CoT) prompting to generate intermediate reasoning chains as the rationale to infer the answer. However, existing CoT studies have primarily focused on the language modality. We propose Multimodal-CoT that incorporates language (text) and vision (images) modalities into a two-stage framework that separates rationale generation and answer inference. In this way, answer inference can leverage better generated rationales that are based on multimodal information. Experimental results on ScienceQA and A-OKVQA benchmark datasets show the effectiveness of our proposed approach. With Multimodal-CoT, our model under 1 billion parameters achieves state-of-the-art performance on the ScienceQA benchmark. Our analysis indicates that Multimodal-CoT offers the advantages of mitigating hallucination and enhancing convergence speed. Code is publicly available at https://github.com/amazon-science/mm-cot.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":"cc-by-sa","is_open_access":true,"peer_reviewed":false,"categories":["Multimodal Machine Learning Applications"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"doi:10.2196/70014","kind":"review","title":"Evaluating Generative AI in Mental Health: Systematic Review of Capabilities and Limitations","authors":["Liying Wang","Tanmay Bhanushali","Zhuoran Huang","Jingyi Yang","Sukriti Badami","Lisa Hightow‐Weidman"],"author_count":6,"year":2025,"published":"2025-02-18","venue":"JMIR Mental Health","cited_by":73,"doi":"10.2196/70014","arxiv_id":null,"nct_id":null,"source_url":"https://doi.org/10.2196/70014","archive_url":"https://space-kitty.com/papers/doi:10.2196/70014","updated":null,"pmid":"40373033","pmcid":null,"openalex_id":"W4407677752","pdf_url":"https://mental.jmir.org/2025/1/e70014/PDF","abstract":"Background: The global shortage of mental health professionals, exacerbated by increasing mental health needs post COVID-19, has stimulated growing interest in leveraging large language models to address these challenges. objectives: This systematic review aims to evaluate the current capabilities of generative artificial intelligence (GenAI) models in the context of mental health applications. Methods: A comprehensive search across 5 databases yielded 1046 references, of which 8 studies met the inclusion criteria. The included studies were original research with experimental designs (eg, Turing tests, sociocognitive tasks, trials, or qualitative methods); a focus on GenAI models; and explicit measurement of sociocognitive abilities (eg, empathy and emotional awareness), mental health outcomes, and user experience (eg, perceived trust and empathy). Results: The studies, published between 2023 and 2024, primarily evaluated models such as ChatGPT-3.5 and 4.0, Bard, and Claude in tasks such as psychoeducation, diagnosis, emotional awareness, and clinical interventions. Most studies used zero-shot prompting and human evaluators to assess the AI responses, using standardized rating scales or qualitative analysis. However, these methods were often insufficient to fully capture the complexity of GenAI capabilities. The reliance on single-shot prompting techniques, limited comparisons, and task-based assessments isolated from a context may oversimplify GenAI's abilities and overlook the nuances of human-artificial intelligence interaction, especially in clinical applications that require contextual reasoning and cultural sensitivity. The findings suggest that while GenAI models demonstrate strengths in psychoeducation and emotional awareness, their diagnostic accuracy, cultural competence, and ability to engage users emotionally remain limited. Users frequently reported concerns about trustworthiness, accuracy, and the lack of emotional engagement. Conclusions: Future research could use more sophisticated evaluation methods, such as few-shot and chain-of-thought prompting to fully uncover GenAI's potential. Longitudinal studies and broader comparisons with human benchmarks are needed to explore the effects of GenAI-integrated mental health care.","abstract_license":"CC-BY (publisher license, via OpenAlex)","license":"cc-by","is_open_access":true,"peer_reviewed":null,"categories":["Artificial Intelligence in Healthcare and Education"],"details":null,"provenance":"openalex","attribution":"Abstract of \"Evaluating Generative AI in Mental Health: Systematic Review of Capabilities and Limitations\" by Liying Wang, Tanmay Bhanushali, Zhuoran Huang et al., https://doi.org/10.2196/70014, used under its Creative Commons Attribution license (CC-BY); markup removed, otherwise unmodified."},{"id":"arxiv:2403.18344","kind":"article","title":"LC-LLM: Explainable lane-change intention and trajectory predictions with Large Language Models","authors":["Mingxing Peng","Xusen Guo","Xianda Chen","Kehua Chen","Meixin Zhu","Long Chen","Fei–Yue Wang"],"author_count":7,"year":2025,"published":"2025-04-03","venue":"Communications in Transportation Research","cited_by":70,"doi":"10.1016/j.commtr.2025.100170","arxiv_id":"2403.18344","nct_id":null,"source_url":"https://doi.org/10.1016/j.commtr.2025.100170","archive_url":"https://space-kitty.com/papers/arxiv:2403.18344","updated":"2024-08-05","pmid":null,"pmcid":null,"openalex_id":"W4409130655","pdf_url":"https://arxiv.org/pdf/2403.18344","abstract":"To ensure safe driving in dynamic environments, autonomous vehicles should possess the capability to accurately predict lane change intentions of surrounding vehicles in advance and forecast their future trajectories. Existing motion prediction approaches have ample room for improvement, particularly in terms of long-term prediction accuracy and interpretability. In this paper, we address these challenges by proposing LC-LLM, an explainable lane change prediction model that leverages the strong reasoning capabilities and self-explanation abilities of Large Language Models (LLMs). Essentially, we reformulate the lane change prediction task as a language modeling problem, processing heterogeneous driving scenario information as natural language prompts for LLMs and employing supervised fine-tuning to tailor LLMs specifically for lane change prediction task. Additionally, we finetune the Chain-of-Thought (CoT) reasoning to improve prediction transparency and reliability, and include explanatory requirements in the prompts during inference stage. Therefore, our LC-LLM model not only predicts lane change intentions and trajectories but also provides CoT reasoning and explanations for its predictions, enhancing its interpretability. Extensive experiments based on the large-scale highD dataset demonstrate the superior performance and interpretability of our LC-LLM in lane change prediction task. To the best of our knowledge, this is the first attempt to utilize LLMs for predicting lane change behavior. Our study shows that LLMs can effectively encode comprehensive interaction information for driving behavior understanding.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":"cc-by","is_open_access":true,"peer_reviewed":null,"categories":["Autonomous Vehicle Technology and Safety"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"doi:10.1109/tsmc.2024.3392930","kind":"article","title":"LLMScenario: Large Language Model Driven Scenario Generation","authors":["Cheng Chang","Siqi Wang","Jiawei Zhang","Jingwei Ge","Li Li"],"author_count":5,"year":2024,"published":"2024-05-13","venue":"IEEE Transactions on Systems Man and Cybernetics Systems","cited_by":83,"doi":"10.1109/tsmc.2024.3392930","arxiv_id":null,"nct_id":null,"source_url":"https://doi.org/10.1109/tsmc.2024.3392930","archive_url":"https://space-kitty.com/papers/doi:10.1109/tsmc.2024.3392930","updated":null,"pmid":null,"pmcid":null,"openalex_id":"W4396877721","pdf_url":null,"abstract":null,"abstract_license":null,"license":null,"is_open_access":false,"peer_reviewed":null,"categories":["Autonomous Vehicle Technology and Safety"],"details":null,"provenance":"openalex","attribution":"Metadata and citation counts: OpenAlex (CC0)."},{"id":"arxiv:2304.02496","kind":"article","title":"Evaluating the ChatGPT family of models for biomedical reasoning and classification","authors":["Shan Chen","Yingya Li","Sheng Fu Lu","Hoang Van","Hugo J.W.L. Aerts","Guergana Savova","Danielle S. Bitterman"],"author_count":7,"year":2023,"published":"2023-12-19","venue":"Journal of the American Medical Informatics Association","cited_by":91,"doi":"10.1093/jamia/ocad256","arxiv_id":"2304.02496","nct_id":null,"source_url":"https://doi.org/10.1093/jamia/ocad256","archive_url":"https://space-kitty.com/papers/arxiv:2304.02496","updated":"2023-04-05","pmid":"38261400","pmcid":null,"openalex_id":"W4391126287","pdf_url":"https://academic.oup.com/jamia/advance-article-pdf/doi/10.1093/jamia/ocad256/56334268/ocad256.pdf","abstract":"Recent advances in large language models (LLMs) have shown impressive ability in biomedical question-answering, but have not been adequately investigated for more specific biomedical applications. This study investigates the performance of LLMs such as the ChatGPT family of models (GPT-3.5s, GPT-4) in biomedical tasks beyond question-answering. Because no patient data can be passed to the OpenAI API public interface, we evaluated model performance with over 10000 samples as proxies for two fundamental tasks in the clinical domain - classification and reasoning. The first task is classifying whether statements of clinical and policy recommendations in scientific literature constitute health advice. The second task is causal relation detection from the biomedical literature. We compared LLMs with simpler models, such as bag-of-words (BoW) with logistic regression, and fine-tuned BioBERT models. Despite the excitement around viral ChatGPT, we found that fine-tuning for two fundamental NLP tasks remained the best strategy. The simple BoW model performed on par with the most complex LLM prompting. Prompt engineering required significant investment.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":true,"peer_reviewed":null,"categories":["Topic Modeling"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"doi:10.1056/aidbp2500120","kind":"article","title":"Assessment of Large Language Models in Clinical Reasoning: A Novel Benchmarking Study","authors":["Liam G. McCoy","Rajiv Swamy","Nidhish Sagar","Minjia Wang","Stephen Bacchi","Jie Ming Nigel Fong","Nigel CK Tan","Kevin Tan","Thomas A. Buckley","Peter G. Brodeur","Leo Anthony Celi","Arjun Kumar Manrai","Aloysius J. Humbert","Adam M. Rodman"],"author_count":14,"year":2025,"published":"2025-09-25","venue":"NEJM AI","cited_by":65,"doi":"10.1056/aidbp2500120","arxiv_id":null,"nct_id":null,"source_url":"https://doi.org/10.1056/aidbp2500120","archive_url":"https://space-kitty.com/papers/doi:10.1056/aidbp2500120","updated":null,"pmid":"42932606","pmcid":null,"openalex_id":"W4414499338","pdf_url":null,"abstract":null,"abstract_license":null,"license":null,"is_open_access":false,"peer_reviewed":null,"categories":["Artificial Intelligence in Healthcare and Education"],"details":null,"provenance":"openalex","attribution":"Metadata and citation counts: OpenAlex (CC0)."},{"id":"arxiv:2210.00720","kind":"preprint","title":"Complexity-Based Prompting for Multi-Step Reasoning","authors":["Yao Fu","Hao Ming Peng","Ashish Sabharwal","Peter A. Clark","Tushar Khot"],"author_count":5,"year":2022,"published":"2022-10-03","venue":null,"cited_by":97,"doi":null,"arxiv_id":"2210.00720","nct_id":null,"source_url":"http://arxiv.org/abs/2210.00720","archive_url":"https://space-kitty.com/papers/arxiv:2210.00720","updated":"2023-01-30","pmid":null,"pmcid":null,"openalex_id":"W4302011807","pdf_url":"https://arxiv.org/pdf/2210.00720","abstract":"We study the task of prompting large-scale language models to perform multi-step reasoning. Existing work shows that when prompted with a chain of thoughts (CoT), sequences of short sentences describing intermediate reasoning steps towards a final answer, large language models can generate new reasoning chains and predict answers for new inputs. A central question is which reasoning examples make the most effective prompts. In this work, we propose complexity-based prompting, a simple and effective example selection scheme for multi-step reasoning. We show that prompts with higher reasoning complexity, i.e., chains with more reasoning steps, achieve substantially better performance on multi-step reasoning tasks over strong baselines. We further extend our complexity-based criteria from prompting (selecting inputs) to decoding (selecting outputs), where we sample multiple reasoning chains from the model, then choose the majority of generated answers from complex reasoning chains (over simple chains). When used to prompt GPT-3 and Codex, our approach substantially improves multi-step reasoning accuracy and achieves new state-of-the-art (SOTA) performance on three math benchmarks (GSM8K, MultiArith, and MathQA) and two BigBenchHard tasks (Date Understanding and Penguins), with an average +5.3 and up to +18 accuracy improvements. Compared with existing example selection schemes like manual tuning or retrieval-based selection, selection based on reasoning complexity is intuitive, easy to implement, and annotation-efficient. Further results demonstrate the robustness of performance gains from complex prompts under format perturbation and distribution shift.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":true,"peer_reviewed":false,"categories":["Topic Modeling"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2309.17447","kind":"article","title":"A Large Language Model Approach to Educational Survey Feedback Analysis","authors":["Michael J. Parker","Caitlin Anderson","Claire Stone","YeaRim Oh"],"author_count":4,"year":2024,"published":"2024-06-25","venue":"International Journal of Artificial Intelligence in Education","cited_by":73,"doi":"10.1007/s40593-024-00414-0","arxiv_id":"2309.17447","nct_id":null,"source_url":"https://doi.org/10.1007/s40593-024-00414-0","archive_url":"https://space-kitty.com/papers/arxiv:2309.17447","updated":"2024-06-27","pmid":null,"pmcid":null,"openalex_id":"W4399995547","pdf_url":"https://arxiv.org/pdf/2309.17447","abstract":"This paper assesses the potential for the large language models (LLMs) GPT-4 and GPT-3.5 to aid in deriving insight from education feedback surveys. Exploration of LLM use cases in education has focused on teaching and learning, with less exploration of capabilities in education feedback analysis. Survey analysis in education involves goals such as finding gaps in curricula or evaluating teachers, often requiring time-consuming manual processing of textual responses. LLMs have the potential to provide a flexible means of achieving these goals without specialized machine learning models or fine-tuning. We demonstrate a versatile approach to such goals by treating them as sequences of natural language processing (NLP) tasks including classification (multi-label, multi-class, and binary), extraction, thematic analysis, and sentiment analysis, each performed by LLM. We apply these workflows to a real-world dataset of 2500 end-of-course survey comments from biomedical science courses, and evaluate a zero-shot approach (i.e., requiring no examples or labeled training data) across all tasks, reflecting education settings, where labeled data is often scarce. By applying effective prompting practices, we achieve human-level performance on multiple tasks with GPT-4, enabling workflows necessary to achieve typical goals. We also show the potential of inspecting LLMs' chain-of-thought (CoT) reasoning for providing insight that may foster confidence in practice. Moreover, this study features development of a versatile set of classification categories, suitable for various course types (online, hybrid, or in-person) and amenable to customization. Our results suggest that LLMs can be used to derive a range of insights from survey text.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":"cc-by","is_open_access":true,"peer_reviewed":null,"categories":["Sentiment Analysis and Opinion Mining"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2506.05566","kind":"article","title":"ScaleRTL: Scaling LLMs with Reasoning Data and Test-Time Compute for Accurate RTL Code Generation","authors":["Chenhui Deng","Yun-Da Tsai","Guan-Ting Liu","Zhongzhi Yu","Haoxing Ren"],"author_count":5,"year":2025,"published":"2025-06-05","venue":"ACM/IEEE Workshop on Machine Learning for CAD (MLCAD)","cited_by":10,"doi":"10.1109/mlcad65511.2025.11189212","arxiv_id":"2506.05566","nct_id":null,"source_url":"https://arxiv.org/abs/2506.05566","archive_url":"https://space-kitty.com/papers/arxiv:2506.05566","updated":"2025-07-15","pmid":null,"pmcid":null,"openalex_id":"W4415003690","pdf_url":"https://arxiv.org/pdf/2506.05566","abstract":"Recent advances in large language models (LLMs) have enabled near-human performance on software coding benchmarks, but their effectiveness in RTL code generation remains limited due to the scarcity of high-quality training data. While prior efforts have fine-tuned LLMs for RTL tasks, they do not fundamentally overcome the data bottleneck and lack support for test-time scaling due to their non-reasoning nature. In this work, we introduce ScaleRTL, the first reasoning LLM for RTL coding that scales up both high-quality reasoning data and test-time compute. Specifically, we curate a diverse set of long chain-of-thought reasoning traces averaging 56K tokens each, resulting in a dataset of 3.5B tokens that captures rich RTL knowledge. Fine-tuning a general-purpose reasoning model on this corpus yields ScaleRTL that is capable of deep RTL reasoning. Subsequently, we further enhance the performance of ScaleRTL through a novel test-time scaling strategy that extends the reasoning process via iteratively reflecting on and self-correcting previous reasoning steps. Experimental results show that ScaleRTL achieves state-of-the-art performance on VerilogEval and RTLLM, outperforming 18 competitive baselines by up to 18.4% on VerilogEval and 12.7% on RTLLM.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":false,"peer_reviewed":null,"categories":["cs.AR","cs.AI"],"details":null,"provenance":"arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."},{"id":"arxiv:2205.05131","kind":"preprint","title":"UL2: Unifying Language Learning Paradigms","authors":["Yi Tay","Mostafa Dehghani","Vinh Q. Tran","Garcia, Xavier","Wei, Jason","Xuezhi Wang","Hyung Won Chung","Shakeri, Siamak","Dara Bahri","Tal Schuster","Huaixiu Zheng","Denny Zhou","Neil Houlsby","Donald Metzler"],"author_count":14,"year":2022,"published":"2022-05-10","venue":null,"cited_by":89,"doi":null,"arxiv_id":"2205.05131","nct_id":null,"source_url":"http://arxiv.org/abs/2205.05131","archive_url":"https://space-kitty.com/papers/arxiv:2205.05131","updated":"2023-02-28","pmid":null,"pmcid":null,"openalex_id":"W4304697829","pdf_url":"https://arxiv.org/pdf/2205.05131","abstract":"Existing pre-trained models are generally geared towards a particular class of problems. To date, there seems to be still no consensus on what the right architecture and pre-training setup should be. This paper presents a unified framework for pre-training models that are universally effective across datasets and setups. We begin by disentangling architectural archetypes with pre-training objectives -- two concepts that are commonly conflated. Next, we present a generalized & unified perspective for self-supervision in NLP and show how different pre-training objectives can be cast as one another and how interpolating between different objectives can be effective. We then propose Mixture-of-Denoisers (MoD), a pre-training objective that combines diverse pre-training paradigms together. We furthermore introduce a notion of mode switching, wherein downstream fine-tuning is associated with specific pre-training schemes. We conduct extensive ablative experiments to compare multiple pre-training objectives and find that our method pushes the Pareto-frontier by outperforming T5 & GPT-like models across multiple diverse setups. By scaling our model up to 20B parameters, we achieve SOTA performance on 50 well-established supervised finetuning based NLP tasks. Our model also achieve strong results at in-context learning, outperforming 175B GPT-3 on zero-shot SuperGLUE and tripling the performance of T5-XXL on one-shot summarization. On 0-shot MMLU, UL2 20B outperforms T0 and T5 models. UL2 20B also works well with chain-of-thought prompting and reasoning, making it an appealing choice for research into reasoning at a small to medium scale of 20B parameters. Finally, we apply FLAN instruction tuning to the UL2 20B model, achieving MMLU and Big-Bench scores competitive to FLAN-PaLM 62B. We release Flax-based T5X checkpoints for the UL2 20B & Flan-UL2 20B.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":true,"peer_reviewed":false,"categories":["Topic Modeling"],"details":null,"provenance":"openalex+arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv."}],"next":"https://space-kitty.com/api/v1/topics/reasoning-models/papers?sort=top&limit=25&offset=25"}