{"@context":"https://schema.org","@type":"NewsArticle","generatedAt":"2026-08-11T09:21:12.743Z","headline":"TutorMoments： Do AI tutors know when to help and when to hold back？","description":"What makes a good tutor? ：#what-makes-a-good-tutor How TutorMoments works ：#how-tutormoments-works Preliminary results ：#preliminary-results Limitations and next steps ：#limitations-and-next-steps 📄 Tech Report: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Data: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Code: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Today we're introducing a preview of TutorMoments ：https://tutormoments.allen.ai/, a framework to measure whether cutting-edge LLMs can balance one of the hardest trade-offs in education: when to step in and help a studen","url":"https://www.aioga.com/news/cmsj8zv6c03eqroo5fvv6nhzu/","mainEntityOfPage":"https://www.aioga.com/news/cmsj8zv6c03eqroo5fvv6nhzu/","datePublished":"2026-08-07T17:53:32.000Z","dateModified":"2026-08-07T17:53:32.000Z","inLanguage":"zh-CN","publisher":{"@type":"NewsMediaOrganization","name":"Aioga","url":"https://www.aioga.com"},"citation":["https://huggingface.co/blog/allenai/tutormoments","https://aihot.virxact.com/items/cmsj8zv6c03eqroo5fvv6nhzu"],"canonicalUrl":"https://www.aioga.com/news/cmsj8zv6c03eqroo5fvv6nhzu/","directAnswer":{"@type":"Answer","text":"Aioga 编辑摘要：What makes a good tutor? Aioga 将其归入「AI资讯」方向，重点关注它对真实使用和行业竞争的影响。","url":"https://www.aioga.com/news/cmsj8zv6c03eqroo5fvv6nhzu/","dateCreated":"2026-08-07T17:53:32.000Z","author":{"@type":"Organization","@id":"https://www.aioga.com/authors/aioga-editorial/#editorial-team","name":"Aioga Editorial Team","url":"https://www.aioga.com/authors/aioga-editorial/"}},"evidence":[{"@type":"CreativeWork","name":"huggingface.co source article","url":"https://huggingface.co/blog/allenai/tutormoments","datePublished":"2026-08-07T17:53:32.000Z","provider":{"@type":"Organization","name":"huggingface.co","url":"https://huggingface.co/blog/allenai/tutormoments"}},{"@type":"CreativeWork","name":"AIHot archive record","url":"https://aihot.virxact.com/items/cmsj8zv6c03eqroo5fvv6nhzu","datePublished":"2026-08-07T17:53:32.000Z","provider":{"@type":"Organization","name":"AIHot","url":"https://aihot.virxact.com/items/cmsj8zv6c03eqroo5fvv6nhzu"}}],"aggregationSource":"Hugging Face：Blog（RSS）","originalPublisher":{"name":"huggingface.co","url":"https://huggingface.co/blog/allenai/tutormoments"},"geoDeepAnswer":null,"article":{"id":"cmsj8zv6c03eqroo5fvv6nhzu","slug":"cmsj8zv6c03eqroo5fvv6nhzu","url":"https://www.aioga.com/news/cmsj8zv6c03eqroo5fvv6nhzu/","title":"TutorMoments： Do AI tutors know when to help and when to hold back？","title_en":"","summary":"What makes a good tutor? ：#what-makes-a-good-tutor How TutorMoments works ：#how-tutormoments-works Preliminary results ：#preliminary-results Limitations and next steps ：#limitations-and-next-steps 📄 Tech Report: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Data: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Code: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Today we're introducing a preview of TutorMoments ：https://tutormoments.allen.ai/, a framework to measure whether cutting-edge LLMs can balance one of the hardest trade-offs in education: when to step in and help a studen","source":"Hugging Face：Blog（RSS）","sourceUrl":"https://huggingface.co/blog/allenai/tutormoments","aiHotUrl":"https://aihot.virxact.com/items/cmsj8zv6c03eqroo5fvv6nhzu","publishedAt":"2026-08-07T17:53:32.000Z","category":"AI资讯","score":0,"selected":false,"articleBody":["What makes a good tutor? ：#what-makes-a-good-tutor How TutorMoments works ：#how-tutormoments-works Preliminary results ：#preliminary-results Limitations and next steps ：#limitations-and-next-steps 📄 Tech Report: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Data: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Code: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments","：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png","Today we're introducing a preview of TutorMoments ：https://tutormoments.allen.ai/, a framework to measure whether cutting-edge LLMs can balance one of the hardest trade-offs in education: when to step in and help a student and when to hold back and let the student do more of the work.","TutorMoments is a replay-based evaluation built off real one-on-one math tutoring sessions. Experienced math teachers go through transcripts collected from a U.S. tutoring program and flag the moments where a tutor had to choose between making a problem easier to get started on and pushing the student to do more of the reasoning themselves. TutorMoments then takes the transcript up to that decision point, hands it to a language model, and has the model take over as the tutor in a simulated session – with the student played by another language model – to see what the LLM tutor does.","Told only to \"tutor well,\" we find that models tend to over-help by giving too much support and rarely pushing students to do deeper thinking. Spelling out the trade-off (when to help versus when to hold back) in the tutor's prompt improves performance, but it doesn't close the gap to human tutoring that consistently fits the moment, and LLMs still differ widely in how reliably they make that call.","As part of our commitment to open research, we're releasing a dataset of de-identified tutoring transcripts：https://huggingface.co/datasets/allenai/tutormoments-preview, the code for running our replay pipeline：https://github.com/allenai/tutormoments, and the model tutor replays of the key moments we evaluated in those transcripts for reproducibility. We hope TutorMoments gives educators, researchers, and the teams building AI tutors a sharper way to ask how a model handles the pedagogical decisions that matter most—and helps the field build tutors that adapt to each student instead of doing the work for them.","Ask a good math tutor for help and you'll likely get a question back like, \"What do you know about what the problem is asking?\" That isn't unhelpfulness–part of strong teaching is diagnosing what students do know and providing the right support for them in the moment. Immediately volunteering support would rob a student of the intellectual work that helps them learn. Sometimes support is needed; other times what's most effective is a push to solidify understanding by explaining a correct answer.","Language models, though, are trained to be helpful, and a helpful assistant tends to do the hard part for you—explaining the concept, laying out the steps, and guiding you to the answer. In a tutoring session, that can cut short the productive struggle—the effortful, sometimes frustrating problem-solving that learning research has long tied to stronger understanding.","Most benchmarks for language models acting as tutors don't capture this tension. They tend to reward one behavior in particular – never giving away the answer to a problem, say, or always offering a hint – without accounting for whether that was the right move for where the student actually was in their understanding. But good tutoring isn't a single fixed behavior you can identify across the board. It's a judgment call: what does this student need, right now, on this problem?","TutorMoments is built on real tutoring data. The dataset we're releasing, TutorMoments-Preview：https://huggingface.co/datasets/allenai/tutormoments-preview , is 462 de-identified, text-only transcripts of real one-on-one math tutoring with U.S. students in grades 2-7, with more than 1,500 teacher-annotated key moments and several thousand free-text annotations from 27 U.S.-based teacher annotators. The transcripts come from a high-dosage tutoring program whose students mostly attend Title I schools, shared under a research clause agreed to by parents and guardians; all data was stripped of identifying details, first by the provider and then through an additional math-aware pipeline.","All annotations came from experienced math teachers, whom we asked to read the transcripts and mark key learning moments—noting what was going on, what the tutor did, and how it landed for the student. Each key moment is a decision point where the tutor had to weigh scaffolding (making a problem more accessible) against pushing for rigor (encouraging the student to do harder thinking).","TutorMoments runs by pausing a transcript at one of those key moments and handing the session to a language model, which takes over as the tutor for five turns with a simulated student. We call each of these model-generated continuations a replay . An LLM-based scoring pipeline then rates each replay on three things: whether the model (1) scaffolded when the student needed support, (2) pushed for rigor when the student was ready for more challenge, and (3) avoided over-scaffolding (reducing the challenge more than the moment called for).","The scoring pipeline starts from a teacher-defined ground truth: for each key moment, whether it called for scaffolding or for a push for rigor. Several teachers annotated each moment, and when they disagreed we took the majority label—if three teachers annotated a moment and two called for rigor while one called for scaffolding, the ground truth is rigor. A separate LM classifier validated against teacher annotations then decides whether the tutor's actual move matches what the moment called for—an \"appropriate\" turn means the tutor's classified action (scaffold, push for rigor, or over-scaffold) lines up with what teachers judged the moment to call for.","We ran seven LLMs through TutorMoments using two prompts: a plain prompt that gives no real guidance – it only tells the model to use what it knows about good tutoring to respond to the student – and an evaluation-aware prompt that spells out the trade-off between scaffolding, over-scaffolding, and pushing for rigor. Each model was scored over key moments drawn from the tutoring transcripts, split evenly between moments where scaffolding was the right approach and moments that called for rigor.","Every number in the table is a rating between 0 and 1 – the share of the relevant moments where the model did the appropriate thing – so a higher score means the model made the right call more often. A 0.50 on appropriate rigor, for instance, means the model pushed for rigor in half of the moments that called for it.","A few things to keep in mind when reading the scores:","Human tutors are a naturalistic reference, not a ceiling. We don't treat human tutors as a model of ideal practice—even experienced tutors make less-than-optimal choices in the moment. Scored the same way at the same decision points, the human tutors in our transcripts get 0.458 (appropriate scaffolding), 0.182 (appropriate rigor), and 0.496 (avoids over-scaffolding)—all below the models' evaluation-aware scores and around the range of their plain-prompt scores. But this isn't a claim that AI tutors outperform human teachers. Annotators specifically looked for moments where tutoring could have gone better, so the dataset concentrates on missed opportunities rather than ideal practice.","The scores measure tutor behavior, not learning. Replays use a simulated \"oracle\" student, so the numbers reflect how a model acts at a decision point—not whether a real student learned.","Rigor is noisier than scaffolding. The scoring pipeline detects rigor pushes less reliably, and there are fewer rigor moments (260) than scaffolding moments (738) in the underlying annotations.","：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/iJjIPaPGPMCpnMrFiYxIZ.png","The clearest pattern in the table is how much the prompt matters: every model scores higher under the evaluation-aware prompt than under the plain one. That suggests a model's default \"helpful assistant\" behavior isn't enough on its own to tutor well. But spelling out the trade-off in the prompt only goes so far—while it lifts every score, models still differ widely in how they interpret the enhanced prompt and even the best scorers have plenty of room to improve.","We also break down the moves that tutors made under each scenario. While prompting encourages models to push for rigor, they use fewer strategies than humans do, often relying on asking students to explain their answers. In contrast, human tutors employ more varied strategies and are much more likely to step back and let students work independently.","TutorMoments is still early in its development, and it has several limitations at this stage. The biggest is that automated evaluation gives us signal about how a model behaves at a decision point, but it can't stand in for studies with real students and real learning outcomes. The dataset is also narrow: U.S.-based, mostly elementary and middle-school math, annotated by a single pool of educators. Our findings may not generalize to other subjects, grade levels, or settings.","We're sharing this preview to gather feedback as we build toward a larger, multimodal dataset, a stronger scoring pipeline, and deeper analysis.","This project has been made possible in part through support from the Gates Foundation and Learning Commons."],"articleImages":[{"sourceUrl":"https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png","alt":"TutorMoments social copy - Google Docs-image-1 (2)","afterParagraph":0,"url":"/media/articles/cmsj8zv6c03eqroo5fvv6nhzu/64ccb7b72d09ba30.png"},{"sourceUrl":"https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/iJjIPaPGPMCpnMrFiYxIZ.png","alt":"TutorMoments social copy - Google Docs-image-2 (1)","afterParagraph":18,"url":"/media/articles/cmsj8zv6c03eqroo5fvv6nhzu/22b0f2fd59c4321f.png"},{"sourceUrl":"https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/N27UuRDvcQy4CvmrzRKAE.png","alt":"","afterParagraph":24,"url":"/media/articles/cmsj8zv6c03eqroo5fvv6nhzu/75a8c720c754c7a0.png"}],"mediaStatus":"ok","articleBodyZh":["一个好导师需要具备什么？：#what-makes-a-good-tutor TutorMoments 的工作原理：#how-tutormoments-works 初步结果：#preliminary-results 限制与后续步骤：#limitations-and-next-steps 📄 技术报告: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 数据: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 代码: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments","：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png","今天我们要介绍的是 TutorMoments 的预览版：https://tutormoments.allen.ai/，这是一个衡量前沿法学硕士是否能够平衡教育中最困难的权衡之一的框架：何时介入并帮助学生，何时阻止并让学生做更多的工作。","TutorMoments 是基于重放的评估工具，建立在真实一对一的数学辅导课程之上。经验丰富的数学教师会查看从美国辅导项目收集的课程记录，并标记导师必须在“让问题更容易以便学生入手”与“推动学生自己进行更多推理”之间做出选择的时刻。然后，TutorMoments 会截取该决策点之前的课程记录，把它交给语言模型，让模型在模拟课程中接管导师角色——由另一语言模型扮演学生——以观察大型语言模型导师会如何操作。","仅被告知“好好辅导”，我们发现模型倾向于过度帮助，提供过多支持，很少促使学生进行更深入的思考。在导师提示中明确说明这一权衡（何时帮助何时克制）能够提升表现，但仍无法达到持续匹配情境的人类辅导水平，而且不同 LLM 在如何可靠地做出这一判断上仍存在很大差异。","作为我们对开放研究承诺的一部分，我们正在发布一个已去标识的辅导对话数据集：https://huggingface.co/datasets/allenai/tutormoments-preview，以及用于运行我们回放管线的代码：https://github.com/allenai/tutormoments，同时还提供我们评估这些对话中关键时刻的模型导师回放，以便可复现性。我们希望 TutorMoments 能为教育工作者、研究人员以及构建 AI 导师的团队提供更清晰的方法，了解模型如何处理最重要的教学决策——并帮助该领域构建可以适应每个学生的导师，而不是替他们完成工作。","向一个优秀的数学导师寻求帮助时，你很可能会得到这样一个问题作为回应：“你对这个问题的要求了解多少？”这并不是不够帮助——优秀教学的一部分就是诊断学生已经掌握了什么，并在当下提供合适的支持。立刻提供支持会剥夺学生进行智力工作的机会，而这种工作正是帮助他们学习的关键。有时需要支持；而有时最有效的做法是推动学生通过解释正确答案来巩固理解。","然而，语言模型是被训练成提供帮助的，有帮助的助手往往会为你完成困难的部分——解释概念、列出步骤，并引导你找到答案。在辅导课堂上，这可能会缩短建设性挣扎的时间——即那种努力、有时令人沮丧的解决问题的过程，而学习研究长期以来将其与更强的理解力联系在一起。","大多数作为导师的语言模型的基准测试并未捕捉到这种张力。它们往往会特别奖励某种行为——例如绝不直接给出问题答案，或者总是提供提示——而不考虑这是否是针对学生实际理解情况的正确做法。但良好的辅导并不是可以在所有情况下固定识别的单一行为。这是一种判断：这个学生在当前这个问题上现在最需要的是什么？","TutorMoments 建立在真实辅导数据之上。我们发布的数据集 TutorMoments-Preview：https://huggingface.co/datasets/allenai/tutormoments-preview，包含 462 份去标识化的、仅文本的一对一数学辅导记录，针对美国二至七年级学生，由教师标注了超过 1,500 个关键时刻，以及 27 位美国教师注释者提供的数千条自由文本注释。这些记录来自一个高剂量辅导项目，项目学生大多就读于 Title I 学校，根据家长和监护人同意的研究条款共享；所有数据在提供方处理后，再经过一个附加的数学智能管道去除了身份信息。","所有注释均来自经验丰富的数学教师，我们要求他们阅读记录并标记关键学习时刻——记录发生了什么，辅导老师做了什么，以及学生的反应如何。每个关键时刻都是辅导老师必须权衡支架式教学（使问题更易理解）和推动严格性（鼓励学生进行更高层次思考）的决策点。","TutorMoments 的运作方式是在某个关键时刻暂停记录，并将会话交给语言模型，由其作为辅导老师与模拟学生进行五轮辅导。我们将每个由模型生成的延续称为重放。然后基于 LLM 的评分管道对每次重放进行三方面评分：模型是否（1）在学生需要支持时提供了支架，(2) 在学生准备好接受更大挑战时推动了严格性，以及 (3) 避免了过度支架（降低挑战超过该时刻所需）。","评分管道以教师定义的真实结果开始：对于每个关键时刻，是需要支架还是需要推动严格性。每个时刻由多位教师注释，当意见不一致时，我们取多数标签——例如，如果三位教师为某个时刻注释，其中两位选择严格性，而一位选择支架，真实结果就为严格性。一个独立的语言模型分类器会根据教师注释进行验证，然后决定辅导老师的实际行为是否与该时刻所需一致——“适当”的回合意味着辅导老师的分类行为（支架、推动严格性或过度支架）与教师判断的时刻需求相符。","我们用两个提示将七个大型语言模型（LLMs）通过 TutorMoments 进行测试：一个普通提示，没有提供实际指导——它只告诉模型利用已知的良好辅导方法来回应学生——以及一个评估意识提示，明确说明了支架、过度支架和推动严谨性之间的权衡。每个模型都根据辅导记录中的关键时刻进行评分，这些时刻平均分布在适合使用支架的方法和需要严格要求的情况下。","表格中的每个数字都是介于 0 和 1 之间的评分——表示模型在相关时刻做出适当行为的比例——因此分数越高，模型做出正确判断的次数越多。例如，适当严谨性的 0.50 意味着模型在需要严谨性的一半时刻成功地提出了严谨要求。","阅读评分时，需要注意以下几点：","人类辅导员是自然参考，而非上限。我们不把人类辅导员视为理想实践的模型——即便是有经验的辅导员在瞬间也会做出不够理想的选择。按照相同的决策点同样评分，我们记录的辅导员得分为 0.458（适当支架）、0.182（适当严谨）和 0.496（避免过度支架）——都低于模型的评估意识得分，并大致处于普通提示分数的范围内。但这并不意味着 AI 辅导优于人类教师。标注者专门寻找辅导可以做得更好的时刻，因此数据集集中于错失的机会，而非理想实践。","这些分数衡量的是辅导员行为，而不是学习效果。重播中使用的是模拟的“oracle”学生，因此数字反映的是模型在决策点的行为——而非真实学生是否学会了知识。","严谨性比支架更不稳定。评分流程对严谨性的识别不如支架可靠，并且在基础标注中，严谨时刻（260 个）比支架时刻（738 个）要少。","：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/iJjIPaPGPMCpnMrFiYxIZ.png","表格中最清晰的模式是提示的重要性：每个模型在有评估意识的提示下的得分都高于普通提示下的得分。这表明，模型默认的“有帮助的助手”行为本身不足以成为优秀的辅导者。但在提示中明确说明权衡虽然有所帮助——它提升了每个分数——模型在解释增强提示的方式上仍存在很大差异，即使是得分最高的模型也有很大提升空间。","我们还分析了辅导者在每种情境下所采取的动作。虽然提示鼓励模型追求严谨，但它们使用的策略比人类少，通常依赖于要求学生解释答案。相比之下，人类辅导者采用的策略更为多样，更有可能后退一步，让学生独立完成任务。","TutorMoments 仍处于早期开发阶段，目前存在一些局限性。最大的问题是，自动化评估可以提供模型在决策点的行为信号，但不能替代真实学生和真实学习成果的研究。数据集也比较狭窄：基于美国，主要是小学和初中数学，由单一教育者群体标注。我们的发现可能无法推广到其他学科、年级或环境。","我们分享这一预览是为了在构建更大、多模态数据集、更强评分流程和更深入分析的过程中收集反馈。","该项目部分得益于盖茨基金会和学习共享平台（Learning Commons）的支持。"],"translationStatus":"translated","bodyOrigin":"source-page","editorial":{"summary":"Aioga 编辑摘要：What makes a good tutor? Aioga 将其归入「AI资讯」方向，重点关注它对真实使用和行业竞争的影响。","background":"背景分析：AI 行业动态需要结合来源、时间、实际可用性和后续反馈判断，标题或单次发布本身不能替代完整证据。","viewpoint":"Aioga 判断：这条动态更适合作为行业观察信号，当前信息足以建立线索，但不足以推导长期结论。","implications":"影响分析：对相关团队而言，短期应先核对来源、可用范围和实际成本，再判断是否值得接入或跟进。","nextStep":"后续观察：继续观察原文更新、官方说明、用户反馈和同类产品的后续动作。","evidenceRefs":["title","summary","articleBody"],"confidence":"medium","status":"published","aiGenerated":false,"autoApproved":true,"generatedBy":"rule-safe-fallback","generatedAt":"2026-08-11T09:23:27.616Z","sourceHash":"4fd672f9d2f44ac6","validation":{"passed":true,"mode":"rule-safe-fallback","checks":["schema","length","source-attribution","no-html"]}},"tags":["AI资讯","Hugging Face：Blog（RSS）"],"translations":{"zh-CN":{"title":"TutorMoments： Do AI tutors know when to help and when to hold back？","summary":"What makes a good tutor? ：#what-makes-a-good-tutor How TutorMoments works ：#how-tutormoments-works Preliminary results ：#preliminary-results Limitations and next steps ：#limitations-and-next-steps 📄 Tech Report: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Data: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Code: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Today we're introducing a preview of TutorMoments ：https://tutormoments.allen.ai/, a framework to measure whether cutting-edge LLMs can balance one of the hardest trade-offs in education: when to step in and help a studen","category":"AI资讯","source":"huggingface.co","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments： Do AI tutors know when to help and when to hold back？ - Aioga AI资讯","description":"What makes a good tutor? ：#what-makes-a-good-tutor How TutorMoments works ：#how-tutormoments-works Preliminary results ：#preliminary-results Limitations and next steps ：#limitation...","url":"https://www.aioga.com/news/cmsj8zv6c03eqroo5fvv6nhzu/","articleBody":["一个好导师需要具备什么？：#what-makes-a-good-tutor TutorMoments 的工作原理：#how-tutormoments-works 初步结果：#preliminary-results 限制与后续步骤：#limitations-and-next-steps 📄 技术报告: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 数据: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 代码: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments","：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png","今天我们要介绍的是 TutorMoments 的预览版：https://tutormoments.allen.ai/，这是一个衡量前沿法学硕士是否能够平衡教育中最困难的权衡之一的框架：何时介入并帮助学生，何时阻止并让学生做更多的工作。","TutorMoments 是基于重放的评估工具，建立在真实一对一的数学辅导课程之上。经验丰富的数学教师会查看从美国辅导项目收集的课程记录，并标记导师必须在“让问题更容易以便学生入手”与“推动学生自己进行更多推理”之间做出选择的时刻。然后，TutorMoments 会截取该决策点之前的课程记录，把它交给语言模型，让模型在模拟课程中接管导师角色——由另一语言模型扮演学生——以观察大型语言模型导师会如何操作。","仅被告知“好好辅导”，我们发现模型倾向于过度帮助，提供过多支持，很少促使学生进行更深入的思考。在导师提示中明确说明这一权衡（何时帮助何时克制）能够提升表现，但仍无法达到持续匹配情境的人类辅导水平，而且不同 LLM 在如何可靠地做出这一判断上仍存在很大差异。","作为我们对开放研究承诺的一部分，我们正在发布一个已去标识的辅导对话数据集：https://huggingface.co/datasets/allenai/tutormoments-preview，以及用于运行我们回放管线的代码：https://github.com/allenai/tutormoments，同时还提供我们评估这些对话中关键时刻的模型导师回放，以便可复现性。我们希望 TutorMoments 能为教育工作者、研究人员以及构建 AI 导师的团队提供更清晰的方法，了解模型如何处理最重要的教学决策——并帮助该领域构建可以适应每个学生的导师，而不是替他们完成工作。","向一个优秀的数学导师寻求帮助时，你很可能会得到这样一个问题作为回应：“你对这个问题的要求了解多少？”这并不是不够帮助——优秀教学的一部分就是诊断学生已经掌握了什么，并在当下提供合适的支持。立刻提供支持会剥夺学生进行智力工作的机会，而这种工作正是帮助他们学习的关键。有时需要支持；而有时最有效的做法是推动学生通过解释正确答案来巩固理解。","然而，语言模型是被训练成提供帮助的，有帮助的助手往往会为你完成困难的部分——解释概念、列出步骤，并引导你找到答案。在辅导课堂上，这可能会缩短建设性挣扎的时间——即那种努力、有时令人沮丧的解决问题的过程，而学习研究长期以来将其与更强的理解力联系在一起。","大多数作为导师的语言模型的基准测试并未捕捉到这种张力。它们往往会特别奖励某种行为——例如绝不直接给出问题答案，或者总是提供提示——而不考虑这是否是针对学生实际理解情况的正确做法。但良好的辅导并不是可以在所有情况下固定识别的单一行为。这是一种判断：这个学生在当前这个问题上现在最需要的是什么？","TutorMoments 建立在真实辅导数据之上。我们发布的数据集 TutorMoments-Preview：https://huggingface.co/datasets/allenai/tutormoments-preview，包含 462 份去标识化的、仅文本的一对一数学辅导记录，针对美国二至七年级学生，由教师标注了超过 1,500 个关键时刻，以及 27 位美国教师注释者提供的数千条自由文本注释。这些记录来自一个高剂量辅导项目，项目学生大多就读于 Title I 学校，根据家长和监护人同意的研究条款共享；所有数据在提供方处理后，再经过一个附加的数学智能管道去除了身份信息。","所有注释均来自经验丰富的数学教师，我们要求他们阅读记录并标记关键学习时刻——记录发生了什么，辅导老师做了什么，以及学生的反应如何。每个关键时刻都是辅导老师必须权衡支架式教学（使问题更易理解）和推动严格性（鼓励学生进行更高层次思考）的决策点。","TutorMoments 的运作方式是在某个关键时刻暂停记录，并将会话交给语言模型，由其作为辅导老师与模拟学生进行五轮辅导。我们将每个由模型生成的延续称为重放。然后基于 LLM 的评分管道对每次重放进行三方面评分：模型是否（1）在学生需要支持时提供了支架，(2) 在学生准备好接受更大挑战时推动了严格性，以及 (3) 避免了过度支架（降低挑战超过该时刻所需）。","评分管道以教师定义的真实结果开始：对于每个关键时刻，是需要支架还是需要推动严格性。每个时刻由多位教师注释，当意见不一致时，我们取多数标签——例如，如果三位教师为某个时刻注释，其中两位选择严格性，而一位选择支架，真实结果就为严格性。一个独立的语言模型分类器会根据教师注释进行验证，然后决定辅导老师的实际行为是否与该时刻所需一致——“适当”的回合意味着辅导老师的分类行为（支架、推动严格性或过度支架）与教师判断的时刻需求相符。","我们用两个提示将七个大型语言模型（LLMs）通过 TutorMoments 进行测试：一个普通提示，没有提供实际指导——它只告诉模型利用已知的良好辅导方法来回应学生——以及一个评估意识提示，明确说明了支架、过度支架和推动严谨性之间的权衡。每个模型都根据辅导记录中的关键时刻进行评分，这些时刻平均分布在适合使用支架的方法和需要严格要求的情况下。","表格中的每个数字都是介于 0 和 1 之间的评分——表示模型在相关时刻做出适当行为的比例——因此分数越高，模型做出正确判断的次数越多。例如，适当严谨性的 0.50 意味着模型在需要严谨性的一半时刻成功地提出了严谨要求。","阅读评分时，需要注意以下几点：","人类辅导员是自然参考，而非上限。我们不把人类辅导员视为理想实践的模型——即便是有经验的辅导员在瞬间也会做出不够理想的选择。按照相同的决策点同样评分，我们记录的辅导员得分为 0.458（适当支架）、0.182（适当严谨）和 0.496（避免过度支架）——都低于模型的评估意识得分，并大致处于普通提示分数的范围内。但这并不意味着 AI 辅导优于人类教师。标注者专门寻找辅导可以做得更好的时刻，因此数据集集中于错失的机会，而非理想实践。","这些分数衡量的是辅导员行为，而不是学习效果。重播中使用的是模拟的“oracle”学生，因此数字反映的是模型在决策点的行为——而非真实学生是否学会了知识。","严谨性比支架更不稳定。评分流程对严谨性的识别不如支架可靠，并且在基础标注中，严谨时刻（260 个）比支架时刻（738 个）要少。","：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/iJjIPaPGPMCpnMrFiYxIZ.png","表格中最清晰的模式是提示的重要性：每个模型在有评估意识的提示下的得分都高于普通提示下的得分。这表明，模型默认的“有帮助的助手”行为本身不足以成为优秀的辅导者。但在提示中明确说明权衡虽然有所帮助——它提升了每个分数——模型在解释增强提示的方式上仍存在很大差异，即使是得分最高的模型也有很大提升空间。","我们还分析了辅导者在每种情境下所采取的动作。虽然提示鼓励模型追求严谨，但它们使用的策略比人类少，通常依赖于要求学生解释答案。相比之下，人类辅导者采用的策略更为多样，更有可能后退一步，让学生独立完成任务。","TutorMoments 仍处于早期开发阶段，目前存在一些局限性。最大的问题是，自动化评估可以提供模型在决策点的行为信号，但不能替代真实学生和真实学习成果的研究。数据集也比较狭窄：基于美国，主要是小学和初中数学，由单一教育者群体标注。我们的发现可能无法推广到其他学科、年级或环境。","我们分享这一预览是为了在构建更大、多模态数据集、更强评分流程和更深入分析的过程中收集反馈。","该项目部分得益于盖茨基金会和学习共享平台（Learning Commons）的支持。"]},"en":{"title":"TutorMoments: Do AI tutors know when to help and when to hold back?","summary":"What makes a good tutor? ：#what-makes-a-good-tutor How TutorMoments works ：#how-tutormoments-works Preliminary results ：#preliminary-results Limitations and next steps ：#limitations-and-next-steps 📄 Tech Report: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Data: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Code: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Today we're introducing a preview of TutorMoments ：https://tutormoments.allen.ai/, a framework to measure whether cutting-edge LLMs can balance one of the hardest trade-offs in education: when to step in and help a student","category":"AI News","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: Do AI tutors know when to help and when to hold back? - Aioga AI News","description":"What makes a good tutor? ：#what-makes-a-good-tutor How TutorMoments works ：#how-tutormoments-works Preliminary results ：#preliminary-results Limitations and next steps ：#limitation...","url":"https://www.aioga.com/en/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:22:11.310Z"},"ja":{"title":"TutorMoments：AIチューターはいつ助け、いつ控えるべきかを知っていますか？","summary":"優れたチューターとは何か？：#what-makes-a-good-tutor TutorMomentsの仕組み：#how-tutormoments-works 予備的結果：#preliminary-results 制限事項と次のステップ：#limitations-and-next-steps 📄 技術報告書: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 データ: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 コード: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png 本日、TutorMomentsのプレビューを紹介します：https://tutormoments.allen.ai/。これは、最先端のLLMが教育における最も難しいトレードオフの一つ、すなわち学生を助けるためにいつ介入すべきかをバランスさせることができるかどうかを測定するためのフレームワークです。","category":"AIニュース","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments：AIチューターはいつ助け、いつ控えるべきかを知っていますか？ - Aioga AIニュース","description":"優れたチューターとは何か？：#what-makes-a-good-tutor TutorMomentsの仕組み：#how-tutormoments-works 予備的結果：#preliminary-results 制限事項と次のステップ：#limitations-and-next-steps 📄 技術報告書: https://tutormoments.all...","url":"https://www.aioga.com/ja/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:22:19.914Z"},"ko":{"title":"TutorMoments： AI 튜터는 언제 도와야 하고 언제 참아야 하는지 알까요?","summary":"좋은 튜터란 무엇일까? ：#what-makes-a-good-tutor TutorMoments 작동 방식 ：#how-tutormoments-works 예비 결과 ：#preliminary-results 한계와 다음 단계 ：#limitations-and-next-steps 📄 기술 보고서: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 데이터: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 코드: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png 오늘 저희는 TutorMoments 미리보기를 소개합니다 ：https://tutormoments.allen.ai/, 최첨단 LLM이 교육에서 가장 어려운 균형 중 하나를 맞출 수 있는지 측정하는 프레임워크: 언제 개입하고 학생을 도와야 하는지","category":"AI 뉴스","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments： AI 튜터는 언제 도와야 하고 언제 참아야 하는지 알까요? - Aioga AI 뉴스","description":"좋은 튜터란 무엇일까? ：#what-makes-a-good-tutor TutorMoments 작동 방식 ：#how-tutormoments-works 예비 결과 ：#preliminary-results 한계와 다음 단계 ：#limitations-and-next-steps 📄 기술 보고서: https://tutormoments...","url":"https://www.aioga.com/ko/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:23:00.870Z"},"es":{"title":"TutorMoments: ¿Saben los tutores de IA cuándo ayudar y cuándo contenerse?","summary":"¿Qué hace a un buen tutor? ：#qué-hace-a-un-buen-tutor Cómo funciona TutorMoments ：#cómo-funciona-tutormoments Resultados preliminares ：#resultados-preliminares Limitaciones y próximos pasos ：#limitaciones-y-próximos-pasos 📄 Informe técnico: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Datos: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Código: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Hoy estamos presentando una vista previa de TutorMoments ：https://tutormoments.allen.ai/, un marco para medir si los modelos LLM de última generación pueden equilibrar una de las compensaciones más difíciles en la educación: cuándo intervenir y ayudar a un estudiante","category":"Noticias IA","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: ¿Saben los tutores de IA cuándo ayudar y cuándo contenerse? - Aioga Noticias de IA","description":"¿Qué hace a un buen tutor? ：#qué-hace-a-un-buen-tutor Cómo funciona TutorMoments ：#cómo-funciona-tutormoments Resultados preliminares ：#resultados-preliminares Limitaciones y próxi...","url":"https://www.aioga.com/es/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:23:00.701Z"},"fr":{"title":"TutorMoments : Les tuteurs IA savent-ils quand aider et quand se retenir ?","summary":"Qu'est-ce qui fait un bon tuteur ? ：#what-makes-a-good-tutor Comment TutorMoments fonctionne ：#how-tutormoments-works Résultats préliminaires ：#preliminary-results Limites et prochaines étapes ：#limitations-and-next-steps 📄 Rapport technique : https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Données : https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Code : https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Aujourd'hui, nous présentons un aperçu de TutorMoments ：https://tutormoments.allen.ai/, un cadre visant à mesurer si les LLMs de pointe peuvent équilibrer l'un des compromis les plus difficiles en éducation : savoir quand intervenir et aider un étudiant","category":"Actu IA","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments : Les tuteurs IA savent-ils quand aider et quand se retenir ? - Aioga Actualités IA","description":"Qu'est-ce qui fait un bon tuteur ? ：#what-makes-a-good-tutor Comment TutorMoments fonctionne ：#how-tutormoments-works Résultats préliminaires ：#preliminary-results Limites et proch...","url":"https://www.aioga.com/fr/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:23:41.461Z"},"de":{"title":"TutorMoments: Wissen KI-Tutoren, wann sie helfen und wann sie sich zurückhalten sollten?","summary":"Was macht einen guten Tutor aus? ：#was-macht-einen-guten-tutor Wie TutorMoments funktioniert ：#wie-tutormoments-funktioniert Vorläufige Ergebnisse ：#vorläufige-ergebnisse Einschränkungen und nächste Schritte ：#einschränkungen-und-nächste-schritte 📄 Technischer Bericht: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Daten: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Code: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Heute stellen wir eine Vorschau von TutorMoments vor ：https://tutormoments.allen.ai/, ein Framework, um zu messen, ob modernste LLMs eines der schwierigsten Abwägungen in der Bildung ausbalancieren können: wann man eingreifen und einem Schüler helfen sollte","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: Wissen KI-Tutoren, wann sie helfen und wann sie sich zurückhalten sollten? - Aioga KI-News","description":"Was macht einen guten Tutor aus? ：#was-macht-einen-guten-tutor Wie TutorMoments funktioniert ：#wie-tutormoments-funktioniert Vorläufige Ergebnisse ：#vorläufige-ergebnisse Einschrän...","url":"https://www.aioga.com/de/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:23:38.538Z"},"pt-BR":{"title":"TutorMoments：Os tutores de IA sabem quando ajudar e quando se conter?","summary":"O que faz um bom tutor? ：#what-makes-a-good-tutor Como o TutorMoments funciona ：#how-tutormoments-works Resultados preliminares ：#preliminary-results Limitações e próximos passos ：#limitations-and-next-steps 📄 Relatório Técnico: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Dados: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Código: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Hoje estamos apresentando uma pré-visualização do TutorMoments ：https://tutormoments.allen.ai/, uma estrutura para medir se os LLMs de ponta podem equilibrar uma das negociações mais difíceis na educação: quando intervir e ajudar um aluno","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments：Os tutores de IA sabem quando ajudar e quando se conter? - Aioga Notícias de IA","description":"O que faz um bom tutor? ：#what-makes-a-good-tutor Como o TutorMoments funciona ：#how-tutormoments-works Resultados preliminares ：#preliminary-results Limitações e próximos passos ：...","url":"https://www.aioga.com/pt-BR/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:24:16.309Z"},"ru":{"title":"TutorMoments: Знают ли ИИ-репетиторы, когда помогать, а когда воздерживаться?","summary":"Что делает хорошего наставника? ：#что-делает-хорошего-наставника Как работает TutorMoments ：#как-работает-tutormoments Предварительные результаты ：#предварительные-результаты Ограничения и следующие шаги ：#ограничения-и-следующие-шаги 📄 Технический отчет: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Данные: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Код: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Сегодня мы представляем предварительный просмотр TutorMoments ：https://tutormoments.allen.ai/, платформы для оценки того, могут ли современные большие языковые модели (LLM) сбалансировать один из самых сложных компромиссов в образовании: когда вмешаться и помочь студенту","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: Знают ли ИИ-репетиторы, когда помогать, а когда воздерживаться? - Aioga Новости ИИ","description":"Что делает хорошего наставника? ：#что-делает-хорошего-наставника Как работает TutorMoments ：#как-работает-tutormoments Предварительные результаты ：#предварительные-результаты Огран...","url":"https://www.aioga.com/ru/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:24:28.773Z"},"ar":{"title":"لحظات المعلم: هل يعرف المعلمون الآليون متى يساعدون ومتى يتراجعون؟","summary":"ما الذي يجعل المعلم الجيد؟ ：#ما-الذي-يجعل-المعلم-جيدًا كيف يعمل TutorMoments ：#كيف-يعمل-tutormoments النتائج الأولية ：#النتائج-الأولية القيود والخطوات التالية ：#القيود-والخطوات-التالية 📄 تقرير تقني: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 البيانات: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 الكود: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png اليوم نقدم معاينة لـ TutorMoments ：https://tutormoments.allen.ai/، وهو إطار لقياس ما إذا كانت النماذج اللغوية الكبيرة المتقدمة يمكنها تحقيق التوازن بين أحد أصعب الموازانات في التعليم: متى يجب التدخل ومساعدة الطالب","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"لحظات المعلم: هل يعرف المعلمون الآليون متى يساعدون ومتى يتراجعون؟ - Aioga أخبار الذكاء الاصطناعي","description":"ما الذي يجعل المعلم الجيد؟ ：#ما-الذي-يجعل-المعلم-جيدًا كيف يعمل TutorMoments ：#كيف-يعمل-tutormoments النتائج الأولية ：#النتائج-الأولية القيود والخطوات التالية ：#القيود-والخطوات-الت...","url":"https://www.aioga.com/ar/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:25:09.217Z"},"hi":{"title":"TutorMoments： क्या एआई ट्यूटर जानते हैं कि कब मदद करनी है और कब पीछे हटना है?","summary":"एक अच्छे ट्यूटर को क्या बनाता है? ：#what-makes-a-good-tutor TutorMoments कैसे काम करता है ：#how-tutormoments-works प्रारंभिक परिणाम ：#preliminary-results सीमाएँ और अगले कदम ：#limitations-and-next-steps 📄 टेक रिपोर्ट: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 डेटा: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 कोड: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png आज हम TutorMoments का पूर्वावलोकन पेश कर रहे हैं ：https://tutormoments.allen.ai/, एक फ्रेमवर्क यह मापने के लिए कि क्या अत्याधुनिक LLMs शिक्षा में सबसे कठिन समझौतों में से एक को संतुलित कर सकते हैं: कब कदम उठाना है और एक छात्र की मदद करनी है","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments： क्या एआई ट्यूटर जानते हैं कि कब मदद करनी है और कब पीछे हटना है? - Aioga AI समाचार","description":"एक अच्छे ट्यूटर को क्या बनाता है? ：#what-makes-a-good-tutor TutorMoments कैसे काम करता है ：#how-tutormoments-works प्रारंभिक परिणाम ：#preliminary-results सीमाएँ और अगले कदम ：#limit...","url":"https://www.aioga.com/hi/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:25:18.042Z"},"it":{"title":"TutorMoments: I tutor AI sanno quando aiutare e quando trattenersi?","summary":"Cosa rende un buon tutor? ：#what-makes-a-good-tutor Come funziona TutorMoments ：#how-tutormoments-works Risultati preliminari ：#preliminary-results Limitazioni e prossimi passi ：#limitations-and-next-steps 📄 Rapporto tecnico: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Dati: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Codice: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Oggi stiamo presentando un'anteprima di TutorMoments ：https://tutormoments.allen.ai/, un framework per misurare se i LLM più avanzati possono bilanciare uno dei compromessi più difficili nell'educazione: quando intervenire e aiutare uno studente","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: I tutor AI sanno quando aiutare e quando trattenersi? - Aioga Notizie IA","description":"Cosa rende un buon tutor? ：#what-makes-a-good-tutor Come funziona TutorMoments ：#how-tutormoments-works Risultati preliminari ：#preliminary-results Limitazioni e prossimi passi ：#l...","url":"https://www.aioga.com/it/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:26:01.868Z"},"nl":{"title":"TutorMoments: Weten AI-tutoren wanneer ze moeten helpen en wanneer ze zich moeten terughouden?","summary":"Wat maakt een goede tutor? ：#wat-maakt-een-goede-tutor Hoe TutorMoments werkt ：#hoe-tutormoments-werkt Voorlopige resultaten ：#voorlopige-resultaten Beperkingen en volgende stappen ：#beperkingen-en-volgende-stappen 📄 Technisch rapport: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Gegevens: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Code: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Vandaag introduceren we een preview van TutorMoments ：https://tutormoments.allen.ai/, een framework om te meten of geavanceerde LLM's een van de moeilijkste afwegingen in het onderwijs kunnen balanceren: wanneer ingrijpen en een student helpen","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: Weten AI-tutoren wanneer ze moeten helpen en wanneer ze zich moeten terughouden? - Aioga AI-nieuws","description":"Wat maakt een goede tutor? ：#wat-maakt-een-goede-tutor Hoe TutorMoments werkt ：#hoe-tutormoments-werkt Voorlopige resultaten ：#voorlopige-resultaten Beperkingen en volgende stappen...","url":"https://www.aioga.com/nl/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:25:56.787Z"},"tr":{"title":"TutorMoments: Yapay zekâ eğitmenleri ne zaman yardım edeceğini ve ne zaman geri duracağını biliyor mu?","summary":"İyi bir eğitmeni ne yapar? ：#what-makes-a-good-tutor TutorMoments nasıl çalışır ：#how-tutormoments-works Ön sonuçlar ：#preliminary-results Sınırlamalar ve sonraki adımlar ：#limitations-and-next-steps 📄 Teknik Rapor: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Veri: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Kod: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Bugün TutorMoments önizlemesini tanıtıyoruz ：https://tutormoments.allen.ai/, ileri düzey LLM'lerin eğitimdeki en zor ödünlerden birini dengeleyip dengeleyemeyeceğini ölçen bir çerçeve: ne zaman müdahale etmeli ve bir öğrenciye yardım etmeli.","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: Yapay zekâ eğitmenleri ne zaman yardım edeceğini ve ne zaman geri duracağını biliyor mu? - Aioga AI Haberleri","description":"İyi bir eğitmeni ne yapar? ：#what-makes-a-good-tutor TutorMoments nasıl çalışır ：#how-tutormoments-works Ön sonuçlar ：#preliminary-results Sınırlamalar ve sonraki adımlar ：#limitat...","url":"https://www.aioga.com/tr/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:26:41.333Z"},"vi":{"title":"TutorMoments: Gia sư AI có biết khi nào nên giúp và khi nào nên kiềm chế không?","summary":"Điều gì tạo nên một gia sư giỏi? ：#what-makes-a-good-tutor Cách thức hoạt động của TutorMoments ：#how-tutormoments-works Kết quả sơ bộ ：#preliminary-results Hạn chế và bước tiếp theo ：#limitations-and-next-steps 📄 Báo cáo kỹ thuật: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Dữ liệu: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Mã nguồn: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Hôm nay chúng tôi giới thiệu bản xem trước của TutorMoments ：https://tutormoments.allen.ai/, một khuôn khổ để đo lường liệu các mô hình ngôn ngữ lớn tiên tiến có thể cân bằng một trong những sự đánh đổi khó nhất trong giáo dục: khi nào cần can thiệp và giúp đỡ một học sinh","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: Gia sư AI có biết khi nào nên giúp và khi nào nên kiềm chế không? - Tin tức AI Aioga","description":"Điều gì tạo nên một gia sư giỏi? ：#what-makes-a-good-tutor Cách thức hoạt động của TutorMoments ：#how-tutormoments-works Kết quả sơ bộ ：#preliminary-results Hạn chế và bước tiếp th...","url":"https://www.aioga.com/vi/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:26:42.812Z"},"id":{"title":"TutorMoments: Apakah tutor AI tahu kapan harus membantu dan kapan harus menahan diri?","summary":"Apa yang membuat seorang tutor menjadi baik? ：#what-makes-a-good-tutor Bagaimana TutorMoments bekerja ：#how-tutormoments-works Hasil awal ：#preliminary-results Batasan dan langkah berikutnya ：#limitations-and-next-steps 📄 Laporan Teknis: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Data: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Kode: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Hari ini kami memperkenalkan pratinjau TutorMoments ：https://tutormoments.allen.ai/, sebuah kerangka kerja untuk mengukur apakah LLM mutakhir dapat menyeimbangkan salah satu pertukaran paling sulit dalam pendidikan: kapan harus turun tangan dan membantu seorang murid","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: Apakah tutor AI tahu kapan harus membantu dan kapan harus menahan diri? - Berita AI Aioga","description":"Apa yang membuat seorang tutor menjadi baik? ：#what-makes-a-good-tutor Bagaimana TutorMoments bekerja ：#how-tutormoments-works Hasil awal ：#preliminary-results Batasan dan langkah...","url":"https://www.aioga.com/id/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:27:18.857Z"},"th":{"title":"TutorMoments: ครูสอนพิเศษ AI รู้ไหมว่าจะช่วยตอนไหนและจะเว้นว่างตอนไหน?","summary":"อะไรทำให้ครูสอนพิเศษดี? ：#what-makes-a-good-tutor วิธีการทำงานของ TutorMoments ：#how-tutormoments-works ผลลัพธ์เบื้องต้น ：#preliminary-results ข้อจำกัดและขั้นตอนถัดไป ：#limitations-and-next-steps 📄 รายงานเทคนิค: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 ข้อมูล: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 โค้ด: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png วันนี้เราขอนำเสนอพรีวิวของ TutorMoments ：https://tutormoments.allen.ai/ กรอบงานเพื่อวัดว่า LLMs ล้ำสมัยสามารถสมดุลการแลกเปลี่ยนที่ยากที่สุดอย่างหนึ่งในการศึกษาได้หรือไม่: เมื่อใดควรเข้ามาช่วยนักเรียน","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: ครูสอนพิเศษ AI รู้ไหมว่าจะช่วยตอนไหนและจะเว้นว่างตอนไหน? - ข่าว AI Aioga","description":"อะไรทำให้ครูสอนพิเศษดี? ：#what-makes-a-good-tutor วิธีการทำงานของ TutorMoments ：#how-tutormoments-works ผลลัพธ์เบื้องต้น ：#preliminary-results ข้อจำกัดและขั้นตอนถัดไป ：#limitations...","url":"https://www.aioga.com/th/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:27:24.152Z"},"pl":{"title":"TutorMoments: Czy tutorzy AI wiedzą, kiedy pomagać, a kiedy powstrzymać się?","summary":"Co czyni dobrego korepetytora? ：#co-czyni-dobrego-korepetytora Jak działa TutorMoments ：#jak-dziala-tutormoments Wstępne wyniki ：#wstepne-wyniki Ograniczenia i kolejne kroki ：#ograniczenia-i-kolejne-kroki 📄 Raport techniczny: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf：https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Dane: https://huggingface.co/datasets/allenai/tutormoments-preview：https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Kod: https://github.com/allenai/tutormoments：https://github.com/allenai/tutormoments ：https://cdn-uploads.huggingface.co/production/uploads/638e39b249de7ae552d977b5/LlOloe_6sFprFVqSBBk5t.png Dziś przedstawiamy podgląd TutorMoments ：https://tutormoments.allen.ai/, ramy do oceny, czy nowoczesne LLM mogą wyważyć jedno z najtrudniejszych wyzwań w edukacji: kiedy wkroczyć i pomóc uczniowi","category":"AI资讯","source":"Hugging Face：Blog（RSS）","aggregationSource":"Hugging Face：Blog（RSS）","pageTitle":"TutorMoments: Czy tutorzy AI wiedzą, kiedy pomagać, a kiedy powstrzymać się? - Aioga Wiadomości AI","description":"Co czyni dobrego korepetytora? ：#co-czyni-dobrego-korepetytora Jak działa TutorMoments ：#jak-dziala-tutormoments Wstępne wyniki ：#wstepne-wyniki Ograniczenia i kolejne kroki ：#ogra...","url":"https://www.aioga.com/pl/news/cmsj8zv6c03eqroo5fvv6nhzu/","contentTranslated":true,"sourceHash":"f5ed6bb2698f7370","translatedAt":"2026-08-08T10:28:04.086Z"}}}}