The paper identifies two main challenges in machine learning education: opaque tools that give learners only a superficial grasp of ML and hinder the development of calibrated trust in AI systems. It introduces ICE‑T, a didactic framework that combines intermodal transfer, computational thinking, and explanatory thinking to provide representational richness, graduated process control, and error contextualization. By linking these facets to trust‑calibration literature, the authors argue that trust should be an explicit educational goal and that ICE‑T offers a scalable method to achieve it.
arXiv:2604. 01114v3 Announce Type: replace-cross Abstract: As generative AI systems are integrated into educational settings, students often encounter AI-generated output while working through learning tasks, either by requesting help or through integrated tools.
By Griffin Pitts, Neha Rani, Weedguet Mildort
arXiv:2512. 12413v2 Announce Type: replace Abstract: Generative AI tools are increasingly embedded in everyday work and learning, yet their fluency, opacity, and propensity to hallucinate mean that users must critically evaluate AI outputs rather than accept them at face value.
By Gabriel R. Lau, Wei Yan Low, Louis Tay, Ysabel Guevarra, Dragan Ga\v{s}evi\'c, Andree Hartanto
arXiv:2606. 05983v1 Announce Type: new Abstract: Generative AI makes answers easy and understanding hard, and uncritical use invites cognitive offloading.
By Alexander Apartsin, Yehudit Aperstein
arXiv:2606. 12441v1 Announce Type: cross Abstract: The four dominant learning theories of behaviorism, cognitivism, constructivism, and connectivism show significant conceptual limitations as generative artificial intelligence (AI) proliferates in educational settings.
By Shan Li, Juan Zheng
arXiv:2608. 13760v1 Announce Type: cross Abstract: Which reasoning behaviors are associated with correct answers in reasoning models, and does reasoning-oriented training amplify those behaviors?
By Jean de Dieu Nyandwi, Leena Mathur, Yonatan Bisk, Robert Hawkins, Graham Neubig
The study examined the impact of expert‑verified AI‑generated study materials on first‑year university economics students. In a two‑cohort difference‑in‑differences design, half of the 170 students received podcasts, FAQs, and quiz‑based guides that were produced by a source‑grounded model and checked by a graduate teaching assistant. The intervention produced a 2.34‑mark advantage on a 50‑mark component, reduced the share of marks below the upper‑second classification boundary by 24.7 percentage points, and had the largest gains for students in the bottom quintile. Interviews with 36 students suggested that the verification label encouraged engagement while preserving critical scrutiny of the AI output.
By Canh Thien Dang, An Nguyen
arXiv:2606. 32032v1 Announce Type: cross Abstract: Metacognition is a critical component of intelligence that describes the ability to monitor and regulate one's own cognitive processes.
By Gabrielle Kaili-May Liu, Avi Caciularu, Gal Yona, Idan Szpektor, Arman Cohan
arXiv:2608.21391v1 Announce Type: cross
Abstract: In this research-to-practice paper we present a survey that can be used to assess students' AI knowledge. As the use of artificial intelligence (AI),...
By Aditya Johri, Cory Brozina, Akriti Bagale
arXiv:2610.07800v1 Announce Type: cross
Abstract: Artificial Intelligence (AI) tools are widely used to support decision making in tasks and domains where no immediate performance feedback is availab...
By Eun Jeong Kang (Xianpi), Peter (Xianpi), Duan, Swati Mishra
The paper argues that explainable AI for computer vision has focused too much on developing interpretability methods rather than assessing how interpretable the models themselves are. It proposes a shift toward model-centric evaluation, using existing tools to compare what different models represent and compute, and emphasizes the need to measure whether humans can truly understand these models. The authors review the current toolbox, survey limited model comparison work, draw parallels to systems neuroscience, and outline a future agenda for model-focused XAI.
By Julien Colin, Nuria Oliver, Thomas Serre
The paper introduces TRUST, a token‑level reward model that predicts the correctness of partial chain‑of‑thought (CoT) traces in vision‑language‑action (VLA) policies, enabling monitoring and selective steering of reasoning. On driving and manipulation VLA benchmarks, TRUST improves reasoning accuracy and reduces collision rates and trajectory errors, though its impact on overall task performance varies across tasks. The study defines two evaluation axes—correctability and actionability—to assess when CoT can serve as a runtime safety interface.
By Sathwik Karnik, Joseph JR. Lee, Aryaman Gupta, Somil Bansal