Typical technique in knowledge distillation (KD) is regularizing the learning of a limited capacity model (student) by pushing its responses to match a powerful model's (teacher). Albeit useful especially in the penultimate layer and beyond, its action on student's feature transform is rather implicit, limiting its practice in the intermediate layers. To explicitly embed the teacher's knowledge in feature transform, we propose a learnable KD layer for the student which improves KD with two distinct abilities: i) learning how to leverage the teacher's knowledge, enabling to discard nuisance information, and ii) feeding forward the transferred knowledge deeper. Thus, the student enjoys the teacher's knowledge during the inference besides training. Formally, we repurpose 1x1-BN-ReLU-1x1 convolution block to assign a semantic vector to each local region according to the template (supervised by the teacher) that the corresponding region of the student matches. To facilitate template learning in the intermediate layers, we propose a novel form of supervision based on the teacher's decisions. Through rigorous experimentation, we demonstrate the effectiveness of our approach on 3 popular classification benchmarks. Code is available at: https://github.com/adagorgun/letKD-framework

But it's not just AI chatbots that kids are encountering in schools and in their daily lives. AI is increasingly everywhere--recommending shows to us on Netflix, helping Alexa answer our questions, powering your favorite interactive Snapchat filters and the way you unlock your smartphone. While some students will invariably be more interested in AI than others, understanding the fundamentals of how these systems work is becoming a basic form of literacy--something everyone who finishes high school should know, says Regina Barzilay, a professor at MIT and a faculty lead for AI at the MIT Jameel Clinic. The clinic recently ran a summer program for 51 high school students interested in the use of AI in health care. Kids should be encouraged to be curious about the systems that play an increasingly prevalent role in our lives, she says.

It was a weekday afternoon in the spring when my son's kindergarten teacher got in touch about the ghost teen. During a social-studies unit about families, the teacher reported, my son had regaled his classmates with tales of his eighteen-year-old brother, who picks him up every afternoon at dismissal. I laughed out loud when I received this note, which was sent via ClassDojo, the messaging app used by our public elementary school in Brooklyn. My son has no brother of any age, and yet I could picture this brother immediately--I imagined him, for some reason, as one of the seniors from "Dazed and Confused," leaning against his scuzzy, old Pontiac parked just outside the school gate, a Marlboro Red hanging from his lips, Foghat wafting from the tape deck. But the teacher did not seem amused.

Three major challenges in reinforcement learning are the complex dynamical systems with large state spaces, the costly data acquisition processes, and the deviation of real-world dynamics from the training environment deployment. To overcome these issues, we study distributionally robust Markov decision processes with continuous state spaces under the widely used Kullback-Leibler, chi-square, and total variation uncertainty sets. We propose a model-based approach that utilizes Gaussian Processes and the maximum variance reduction algorithm to efficiently learn multi-output nominal transition dynamics, leveraging access to a generative model (i.e., simulator). We further demonstrate the statistical sample complexity of the proposed method for different uncertainty sets. These complexity bounds are independent of the number of states and extend beyond linear dynamics, ensuring the effectiveness of our approach in identifying near-optimal distributionally-robust policies. The proposed method can be further combined with other model-free distributionally robust reinforcement learning methods to obtain a near-optimal robust policy. Experimental results demonstrate the robustness of our algorithm to distributional shifts and its superior performance in terms of the number of samples needed.

This paper offers a novel conceptual framework comprising four essential cognitive mechanisms that operate concurrently and collaboratively to enable metalearning (knowledge and regulation of learning) strategy implementation in young children. A roadmap incorporating the core mechanisms and the associated strategies is presented as an explanation of the developing brain's remarkable cross-context learning competence. The tetrad of fundamental complementary processes is chosen to collectively represent the bare-bones metalearning architecture that can be extended to artificial intelligence (AI) systems emulating brain-like learning and problem-solving skills. Utilizing the metalearning-enabled young mind as a model for brain-inspired computing, this work further discusses important implications for morally grounded AI.

In the digital era, the integration of artificial intelligence (AI) in education has ushered in transformative changes, redefining teaching methodologies, curriculum planning, and student engagement. This review paper delves deep into the rapidly evolving landscape of digital education by contrasting the capabilities and impact of OpenAI's pioneering text generation tools like Bing Chat, Bard, Ernie with a keen focus on the novel ChatGPT. Grounded in a typology that views education through the lenses of system, process, and result, the paper navigates the multifaceted applications of AI. From decentralizing global education and personalizing curriculums to digitally documenting competence-based outcomes, AI stands at the forefront of educational modernization. Highlighting ChatGPT's meteoric rise to one million users in just five days, the study underscores its role in democratizing education, fostering autodidacticism, and magnifying student engagement. However, with such transformative power comes the potential for misuse, as text-generation tools can inadvertently challenge academic integrity. By juxtaposing the promise and pitfalls of AI in education, this paper advocates for a harmonized synergy between AI tools and the educational community, emphasizing the urgent need for ethical guidelines, pedagogical adaptations, and strategic collaborations.

Source code summarization is the task of writing natural language descriptions of source code. A typical use case is generating short summaries of subroutines for use in API documentation. The heart of almost all current research into code summarization is the encoder-decoder neural architecture, and the encoder input is almost always a single subroutine or other short code snippet. The problem with this setup is that the information needed to describe the code is often not present in the code itself -- that information often resides in other nearby code. In this paper, we revisit the idea of ``file context'' for code summarization. File context is the idea of encoding select information from other subroutines in the same file. We propose a novel modification of the Transformer architecture that is purpose-built to encode file context and demonstrate its improvement over several baselines. We find that file context helps on a subset of challenging examples where traditional approaches struggle.

Recent works usually address Dialog policy learning DPL by training a reinforcement learning (RL) agent to determine the best dialog action. However, existing works on deep RL require a large volume of agent-user interactions to achieve acceptable performance. In this paper, we propose to make full use of the plain text knowledge from the pre-trained language model to accelerate the RL agent's learning speed. Specifically, we design a dialog action-aware transformer encoder (DaTrans), which integrates a new fine-tuning procedure named masked last action task to encourage DaTrans to be dialog-aware and distils action-specific features. Then, DaTrans is further optimized in an RL setting with ongoing interactions and evolves through exploration in the dialog action space toward maximizing long-term accumulated rewards. The effectiveness and efficiency of the proposed model are demonstrated with both simulator evaluation and human evaluation.

Reasoning is a cognitive process of using evidence to reach a sound conclusion. The reasoning capability is essential for large language models (LLMs) to serve as the brain of the artificial general intelligence agent. Recent studies reveal that fine-tuning LLMs on data with the chain of thought (COT) reasoning process can significantly enhance their reasoning capabilities. However, we find that the fine-tuned LLMs suffer from an \textit{Assessment Misalignment} problem, i.e., they frequently assign higher scores to subpar COTs, leading to potential limitations in their reasoning abilities. To address this problem, we introduce an \textit{Alignment Fine-Tuning (AFT)} paradigm, which involves three steps: 1) fine-tuning LLMs with COT training data; 2) generating multiple COT responses for each question, and categorizing them into positive and negative ones based on whether they achieve the correct answer; 3) calibrating the scores of positive and negative responses given by LLMs with a novel constraint alignment loss. Specifically, the constraint alignment loss has two objectives: a) Alignment, which guarantees that positive scores surpass negative scores to encourage answers with high-quality COTs; b) Constraint, which keeps the negative scores confined to a reasonable range to prevent the model degradation. Beyond just the binary positive and negative feedback, the constraint alignment loss can be seamlessly adapted to the ranking situations when ranking feedback is accessible. Furthermore, we also delve deeply into recent ranking-based alignment methods, such as DPO, RRHF, and PRO, and discover that the constraint, which has been overlooked by these approaches, is also crucial for their performance. Extensive experiments on four reasoning benchmarks with both binary and ranking feedback demonstrate the effectiveness of AFT.

This paper explores using generative AI and aesthetics to promote cultural creativity in rural China amidst COVID-19's impact. Through literature reviews, case studies, surveys, and text analysis, it examines art and technology applications in rural contexts and identifies key challenges. The study finds artworks often fail to resonate locally, while reliance on external artists limits sustainability. Hence, nurturing grassroots "artist villagers" through AI is proposed. Our approach involves training machine learning on subjective aesthetics to generate culturally relevant content. Interactive AI media can also boost tourism while preserving heritage. This pioneering research puts forth original perspectives on the intersection of AI and aesthetics to invigorate rural culture. It advocates holistic integration of technology and emphasizes AI's potential as a creative enabler versus replacement. Ultimately, it lays the groundwork for further exploration of leveraging AI innovations to empower rural communities. This timely study contributes to growing interest in emerging technologies to address critical issues facing rural China.