Nov 2026· COMMUNITY Jurnal Pengabdian Kepada Masyarakat· 7 references
Artificial Intelligence in Education
Abstract
The utilization of Generative Artificial Intelligence (GenAI), specifically ChatGPT, holds strong potential for accelerating the design of instructional materials in schools. This qualitative descriptive study aims to describe the implementation of ChatGPT training for teachers at SMP Negeri 40 Samarinda, identify the technical challenges encountered, and analyze changes in participants' understanding when drafting teaching modules and Student Worksheets (LKPD). The training involved 24 teachers, from whom five participants were purposively selected as in-depth interview informants based on subject diversity and active workshop participation. Data were gathered through observational procedures, artifact analysis of generated instructional designs, and post-training interviews analyzed thematically. The findings indicate that the training offered structured experiences for participants in formulating prompts based on Learning Outcomes (Capaian Pembelajaran) and Learning Objectives (Tujuan Pembelajaran). The primary challenges identified centered on fine-tuning prompt precision, limited simulation time, and presentation screen visibility. Based on artifact comparisons between initial and revised drafts, participants demonstrated critical evaluation skills by correcting contextual discrepancies in ChatGPT outputs before classroom integration. ABSTRAK Pemanfaatan Generative Artificial Intelligence (GenAI), khususnya ChatGPT, berpotensi mempercepat penyusunan perangkat pembelajaran di sekolah. Penelitian deskriptif kualitatif ini bertujuan untuk mendeskripsikan pelaksanaan pelatihan ChatGPT bagi guru di SMP Negeri 40 Samarinda, mengidentifikasi kendala teknis yang dihadapi, serta menganalisis perubahan pemahaman peserta dalam menyusun modul ajar dan Lembar Kerja Peserta Didik (LKPD). Pelatihan ini diikuti oleh 24 guru, dengan 5 peserta dipilih secara bertujuan (purposive) sebagai informan wawancara mendalam berdasarkan variasi mata pelajaran dan keaktifan praktik. Data dikumpulkan melalui observasi proses, analisis dokumen rancangan pembelajaran, serta wawancara pascapraktik yang dianalisis secara tematik. Hasil penelitian menunjukkan bahwa pelatihan memberikan pengalaman terarah bagi peserta dalam merumuskan instruksi (prompt) berbasis Capaian Pembelajaran (CP) dan Tujuan Pembelajaran (TP). Kendala utama yang teridentifikasi berpusat pada penyesuaian instruksi yang presisi, alokasi waktu simulasi yang terbatas, serta visibilitas media presentasi. Berdasarkan perbandingan draf awal dan akhir, peserta mulai menunjukkan kemampuan evaluasi kritis dengan menyunting kekeliruan kontekstual keluaran ChatGPT sebelum diintegrasikan ke dalam kelas.
This paper investigates developing real-world perception for active agents, proposes Gibson Environment for this purpose, and showcases a set of perceptual tasks learned therein.
F. Xia, Amir Zamir, Zhi-Yang He et al.· 2018 IEEE/CVF Conference on...· 1,069 citations· ⚡112
GAOKAO-Bench is introduced, an intuitive benchmark that employs questions from the Chinese GAOKAO examination as test samples, including both subjective and objective questions that contribute a robust evaluation benchmark for future large language models and offers valuable insights into the advantages and limitations...
Xiaotian Zhang, Chun-yan Li, Yi Zong et al.· arXiv.org· 216 citations· ⚡17
An algorithm is proposed which leverages disentangled semantic factors to generate adversarial perturbation by altering controlled semantic attributes to fool the learner towards various "adversarial" targets.
Haonan Qiu, Chaowei Xiao, Lei Yang et al.· arXiv.org· 208 citations· ⚡17
This work proposes a HALO that directly maximizes the utility of generations instead of maximizing the log-likelihood of preferences, and suggests that the best loss depends on the inductive biases most appropriate for a given setting, an oft-overlooked consideration.
This work investigates the possibilities of using LLMs in a resume screening setting via a document retrieval framework that simulates job candidate selection and finds that the MTEs are biased, significantly favoring White-associated names in 85% of cases and female-associated names in only 11.1% of cases.
This work shows that orders of magnitude enhancement in performance could be obtained by a combination of hardware improvements and tight quantum-HPC integration and introduces high-performance architectures for quantum-probabilistic computing with custom-designed accelerators to tackle today's industry-scale classical...
Masoud Mohseni, Artur Scherer, K. Johnson et al.· arXiv.org· 121 citations· ⚡9