Effat FarhanaSantu, Souvika SarkarSantu, Ralph KnipperSantu, Indrani DeySantu, Hari NarayananSantu, Sadhana PuntambekarSantu, Shubhra Kanti KarmakerSantu
{"title":"SimPal:建立元对话框架,了解 K-12 物理教师的教学目标","authors":"Effat FarhanaSantu, Souvika SarkarSantu, Ralph KnipperSantu, Indrani DeySantu, Hari NarayananSantu, Sadhana PuntambekarSantu, Shubhra Kanti KarmakerSantu","doi":"arxiv-2407.06241","DOIUrl":null,"url":null,"abstract":"Simulations are widely used to teach science in grade schools. These\nsimulations are often augmented with a conversational artificial intelligence\n(AI) agent to provide real-time scaffolding support for students conducting\nexperiments using the simulations. AI agents are highly tailored for each\nsimulation, with a predesigned set of Instructional Goals (IGs), making it\ndifficult for teachers to adjust IGs as the agent may no longer align with the\nrevised IGs. Additionally, teachers are hesitant to adopt new third-party\nsimulations for the same reasons. In this research, we introduce SimPal, a\nLarge Language Model (LLM) based meta-conversational agent, to solve this\nmisalignment issue between a pre-trained conversational AI agent and the\nconstantly evolving pedagogy of instructors. Through natural conversation with\nSimPal, teachers first explain their desired IGs, based on which SimPal\nidentifies a set of relevant physical variables and their relationships to\ncreate symbolic representations of the desired IGs. The symbolic\nrepresentations can then be leveraged to design prompts for the original AI\nagent to yield better alignment with the desired IGs. We empirically evaluated\nSimPal using two LLMs, ChatGPT-3.5 and PaLM 2, on 63 Physics simulations from\nPhET and Golabz. Additionally, we examined the impact of different prompting\ntechniques on LLM's performance by utilizing the TELeR taxonomy to identify\nrelevant physical variables for the IGs. Our findings showed that SimPal can do\nthis task with a high degree of accuracy when provided with a well-defined\nprompt.","PeriodicalId":501565,"journal":{"name":"arXiv - PHYS - Physics Education","volume":"29 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"SimPal: Towards a Meta-Conversational Framework to Understand Teacher's Instructional Goals for K-12 Physics\",\"authors\":\"Effat FarhanaSantu, Souvika SarkarSantu, Ralph KnipperSantu, Indrani DeySantu, Hari NarayananSantu, Sadhana PuntambekarSantu, Shubhra Kanti KarmakerSantu\",\"doi\":\"arxiv-2407.06241\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Simulations are widely used to teach science in grade schools. These\\nsimulations are often augmented with a conversational artificial intelligence\\n(AI) agent to provide real-time scaffolding support for students conducting\\nexperiments using the simulations. AI agents are highly tailored for each\\nsimulation, with a predesigned set of Instructional Goals (IGs), making it\\ndifficult for teachers to adjust IGs as the agent may no longer align with the\\nrevised IGs. Additionally, teachers are hesitant to adopt new third-party\\nsimulations for the same reasons. In this research, we introduce SimPal, a\\nLarge Language Model (LLM) based meta-conversational agent, to solve this\\nmisalignment issue between a pre-trained conversational AI agent and the\\nconstantly evolving pedagogy of instructors. Through natural conversation with\\nSimPal, teachers first explain their desired IGs, based on which SimPal\\nidentifies a set of relevant physical variables and their relationships to\\ncreate symbolic representations of the desired IGs. The symbolic\\nrepresentations can then be leveraged to design prompts for the original AI\\nagent to yield better alignment with the desired IGs. We empirically evaluated\\nSimPal using two LLMs, ChatGPT-3.5 and PaLM 2, on 63 Physics simulations from\\nPhET and Golabz. Additionally, we examined the impact of different prompting\\ntechniques on LLM's performance by utilizing the TELeR taxonomy to identify\\nrelevant physical variables for the IGs. Our findings showed that SimPal can do\\nthis task with a high degree of accuracy when provided with a well-defined\\nprompt.\",\"PeriodicalId\":501565,\"journal\":{\"name\":\"arXiv - PHYS - Physics Education\",\"volume\":\"29 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-07-08\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - PHYS - Physics Education\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2407.06241\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - PHYS - Physics Education","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2407.06241","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
SimPal: Towards a Meta-Conversational Framework to Understand Teacher's Instructional Goals for K-12 Physics
Simulations are widely used to teach science in grade schools. These
simulations are often augmented with a conversational artificial intelligence
(AI) agent to provide real-time scaffolding support for students conducting
experiments using the simulations. AI agents are highly tailored for each
simulation, with a predesigned set of Instructional Goals (IGs), making it
difficult for teachers to adjust IGs as the agent may no longer align with the
revised IGs. Additionally, teachers are hesitant to adopt new third-party
simulations for the same reasons. In this research, we introduce SimPal, a
Large Language Model (LLM) based meta-conversational agent, to solve this
misalignment issue between a pre-trained conversational AI agent and the
constantly evolving pedagogy of instructors. Through natural conversation with
SimPal, teachers first explain their desired IGs, based on which SimPal
identifies a set of relevant physical variables and their relationships to
create symbolic representations of the desired IGs. The symbolic
representations can then be leveraged to design prompts for the original AI
agent to yield better alignment with the desired IGs. We empirically evaluated
SimPal using two LLMs, ChatGPT-3.5 and PaLM 2, on 63 Physics simulations from
PhET and Golabz. Additionally, we examined the impact of different prompting
techniques on LLM's performance by utilizing the TELeR taxonomy to identify
relevant physical variables for the IGs. Our findings showed that SimPal can do
this task with a high degree of accuracy when provided with a well-defined
prompt.