Dein Suchergebnis zum Thema: Model

Neue Folge des IPN-Podcasts "Im Dialog" jetzt online: Large Language Models und ihre

https://www.leibniz-ipn.de/de/das-ipn/aktuelles/aktuelle-meldungen/neue-folge-des-ipn-podcasts-im-dialog-jetzt-online-large-language-models-und-ihre-potenziale-im-bildungssystem

Dialog" diskutieren hochkarätige Expert:innen das Potenzial von Large Language Models – Wissenschaftlichen Diskussion (SWK) kürzlich veröffentlichte Impulspapier „Large Language Models
Meldungen Neue Folge des IPN-Podcasts "Im Dialog" jetzt online: Large Language Models

ASLAN at BEA 2026 Shared Task 2: Voting across scoring paradigms – Leibniz-Institut für die Pädagogik der Naturwissenschaften und Mathematik

https://www.leibniz-ipn.de/de/forschen/publikationen/aslan-at-bea-2026-shared-task-2-voting-across-scoring-paradigms?show_navhelper=1

prompts observed during training, we compare question-specific and generic scoring models – For the unseen questions track, where models must generalize to previously unseen – Our experiments show that similarity-based models outperform instance-based models – and LLMbased models in the unseen answers setting. – In addition, we find that ensemble methods improve robustness over individual models
prompts observed during training, we compare question-specific and generic scoring models

ASLAN at BEA 2026 Shared Task 2: Voting across scoring paradigms – Leibniz-Institut für die Pädagogik der Naturwissenschaften und Mathematik

https://www.leibniz-ipn.de/de/forschen/publikationen/aslan-at-bea-2026-shared-task-2-voting-across-scoring-paradigms

prompts observed during training, we compare question-specific and generic scoring models – For the unseen questions track, where models must generalize to previously unseen – Our experiments show that similarity-based models outperform instance-based models – and LLMbased models in the unseen answers setting. – In addition, we find that ensemble methods improve robustness over individual models
prompts observed during training, we compare question-specific and generic scoring models

Beyond analysing frequencies: Exploring teacher professional vision with epistemic network analysis of teachers’ think-aloud data – Leibniz-Institut für die Pädagogik der Naturwissenschaften und Mathematik

https://www.leibniz-ipn.de/de/forschen/publikationen/beyond-analyzing-frequencies?show_navhelper=1

think-aloud protocols can offer useful insights into novice and expert teachers’ mental models – integrate information and scripts into more elaborate and well-structured mental models – useful novel method to examine teachers’ integration of information into mental models
think-aloud protocols can offer useful insights into novice and expert teachers’ mental models

Beyond analysing frequencies: Exploring teacher professional vision with epistemic network analysis of teachers’ think-aloud data – Leibniz-Institut für die Pädagogik der Naturwissenschaften und Mathematik

https://www.leibniz-ipn.de/de/forschen/publikationen/beyond-analyzing-frequencies

think-aloud protocols can offer useful insights into novice and expert teachers’ mental models – integrate information and scripts into more elaborate and well-structured mental models – useful novel method to examine teachers’ integration of information into mental models
think-aloud protocols can offer useful insights into novice and expert teachers’ mental models

Fifth and sixth graders speak more richly than they write: Benefits of speech input for digitally assessing conceptual understanding – Leibniz-Institut für die Pädagogik der Naturwissenschaften und Mathematik

https://www.leibniz-ipn.de/de/forschen/publikationen/fifth-and-sixth-graders-speak-more-richly-than-they-write

capable of automatic speech recognition (ASR) and the improved performance of ASR models – Results from multi-level models indicated that speech input answers were substantially
capable of automatic speech recognition (ASR) and the improved performance of ASR models

Fifth and sixth graders speak more richly than they write: Benefits of speech input for digitally assessing conceptual understanding – Leibniz-Institut für die Pädagogik der Naturwissenschaften und Mathematik

https://www.leibniz-ipn.de/de/forschen/publikationen/fifth-and-sixth-graders-speak-more-richly-than-they-write?show_navhelper=1

capable of automatic speech recognition (ASR) and the improved performance of ASR models – Results from multi-level models indicated that speech input answers were substantially
capable of automatic speech recognition (ASR) and the improved performance of ASR models