Offline-to-online reinforcement learning (O2O RL) aims to obtain a continually improving policy as it interacts with the environment, while ensuring the initial behaviour is satisficing. This satisficing behaviour is necessary for robotic manipulation where random exploration can be costly due to catastrophic failures and time. O2O RL is especially compelling when we can only obtain a scarce amount of (potentially suboptimal) demonstrations$\unicode{x2014}$a scenario where behavioural cloning (BC) is known to suffer from distribution shift. Previous works have outlined the challenges in applying O2O RL algorithms under the image-based environments. In this work, we propose a novel O2O RL algorithm that can learn in a real-life image-based robotic vacuum grasping task with a small number of demonstrations where BC fails majority of the time. The proposed algorithm replaces the target network in off-policy actor-critic algorithms with a regularization technique inspired by neural tangent kernel. We demonstrate that the proposed algorithm can reach above 90% success rate in under two hours of interaction time, with only 50 human demonstrations, while BC and two commonly-used RL algorithms fail to achieve similar performance.

Guiding users through complex procedural plans is an inherently multimodal task in which having visually illustrated plan steps is crucial to deliver an effective plan guidance. However, existing works on plan-following language models (LMs) often are not capable of multimodal input and output. In this work, we present MM-PlanLLM, the first multimodal LLM designed to assist users in executing instructional tasks by leveraging both textual plans and visual information. Specifically, we bring cross-modality through two key tasks: Conversational Video Moment Retrieval, where the model retrieves relevant step-video segments based on user queries, and Visually-Informed Step Generation, where the model generates the next step in a plan, conditioned on an image of the user's current progress. MM-PlanLLM is trained using a novel multitask-multistage approach, designed to gradually expose the model to multimodal instructional-plans semantic layers, achieving strong performance on both multimodal and textual dialogue in a plan-grounded setting. Furthermore, we show that the model delivers cross-modal temporal and plan-structure representations aligned between textual plan steps and instructional video moments.

There are obvious benefits to integrating generative AI (artificial intelligence) into language learning and teaching. Those include using AI as a language tutor, creating learning materials, or assessing learner output. However, due to how AI systems under-stand human language, based on a mathematical model using statistical probability, they lack the lived experience to be able to use language with the same social aware-ness as humans. Additionally, there are built-in linguistic and cultural biases based on their training data which is mostly in English and predominantly from Western sources. Those facts limit AI suitability for some language learning interactions. Stud-ies have clearly shown that systems such as ChatGPT often do not produce language that is pragmatically appropriate. The lack of linguistic and cultural authenticity has important implications for how AI is integrated into second language acquisition as well as in instruction targeting development of intercultural communication compe-tence.

This paper considers an online reinforcement learning algorithm that leverages pre-collected data (passive memory) from the environment for online interaction. We show that using passive memory improves performance and further provide theoretical guarantees for regret that turns out to be near-minimax optimal. Results show that the quality of passive memory determines sub-optimality of the incurred regret. The proposed approach and results hold in both continuous and discrete state-action spaces.

This paper introduces SELAR, a framework designed to effectively help teachers integrate artificial intelligence (AI) into their curriculum. The framework was designed by running workshops organized to gather lecturers' feedback. In this paper, we assess the effectiveness of the framework through additional workshops organized with lecturers from the Hague University of Applied Sciences. The workshops tested the application of the framework to adapt existing courses to leverage generative AI technology. Each participant was tasked to apply SELAR to one of their learning goals in order to evaluate AI integration potential and, if successful, to update the teaching methods accordingly. Findings show that teachers were able to effectively use the SELAR to integrate generative AI into their courses. Future work will focus on providing additional guidance and examples to use the framework more effectively.

With the increasing adoption of large language models (LLMs) in education, concerns about inherent biases in these models have gained prominence. We evaluate LLMs for bias in the personalized educational setting, specifically focusing on the models' roles as "teachers". We reveal significant biases in how models generate and select educational content tailored to different demographic groups, including race, ethnicity, sex, gender, disability status, income, and national origin. We introduce and apply two bias score metrics--Mean Absolute Bias (MAB) and Maximum Difference Bias (MDB)--to analyze 9 open and closed state-of-the-art LLMs. Our experiments, which utilize over 17,000 educational explanations across multiple difficulty levels and topics, uncover that models perpetuate both typical and inverted harmful stereotypes.

Imitation learning (IL) enables agents to acquire skills directly from expert demonstrations, providing a compelling alternative to reinforcement learning. However, prior online IL approaches struggle with complex tasks characterized by high-dimensional inputs and complex dynamics. In this work, we propose a novel approach to online imitation learning that leverages reward-free world models. Our method learns environmental dynamics entirely in latent spaces without reconstruction, enabling efficient and accurate modeling. We adopt the inverse soft-Q learning objective, reformulating the optimization process in the Q-policy space to mitigate the instability associated with traditional optimization in the reward-policy space. By employing a learned latent dynamics model and planning for control, our approach consistently achieves stable, expert-level performance in tasks with high-dimensional observation or action spaces and intricate dynamics. We evaluate our method on a diverse set of benchmarks, including DMControl, MyoSuite, and ManiSkill2, demonstrating superior empirical performance compared to existing approaches. Imitation learning (IL) has garnered considerable attention due to its broad applications across various domains, such as robotic manipulation (Zhu et al., 2023; Chi et al., 2023) and autonomous driving (Hu et al., 2022; Zhou et al., 2021). Unlike reinforcement learning, where agents learn through reward signals, IL involves learning directly from expert demonstrations.

Large Language Models (LLMs) were shown to struggle with long-term planning, which may be caused by the limited way in which they explore the space of possible solutions. We propose an architecture where a Reinforcement Learning (RL) Agent guides an LLM's space exploration: (1) the Agent has access to domain-specific information, and can therefore make decisions about the quality of candidate solutions based on specific and relevant metrics, which were not explicitly considered by the LLM's training objective; (2) the LLM can focus on generating immediate next steps, without the need for long-term planning. We allow non-linear reasoning by exploring alternative paths and backtracking. We evaluate this architecture on the program equivalence task, and compare it against Chain of Thought (CoT) and Tree of Thoughts (ToT). We assess both the downstream task, denoting the binary classification, and the intermediate reasoning steps. Our approach compares positively against CoT and ToT.

Incremental Few-Shot Semantic Segmentation (iFSS) tackles a task that requires a model to continually expand its segmentation capability on novel classes using only a few annotated examples. Typical incremental approaches encounter a challenge that the objective of the base training phase (fitting base classes with sufficient instances) does not align with the incremental learning phase (rapidly adapting to new classes with less forgetting). This disconnect can result in suboptimal performance in the incremental setting. This study introduces a meta-learning-based prototype approach that encourages the model to learn how to adapt quickly while preserving previous knowledge. Concretely, we mimic the incremental evaluation protocol during the base training session by sampling a sequence of pseudo-incremental tasks. Each task in the simulated sequence is trained using a meta-objective to enable rapid adaptation without forgetting. To enhance discrimination among class prototypes, we introduce prototype space redistribution learning, which dynamically updates class prototypes to establish optimal inter-prototype boundaries within the prototype space. Extensive experiments on iFSS datasets built upon PASCAL and COCO benchmarks show the advanced performance of the proposed approach, offering valuable insights for addressing iFSS challenges.

Text-to-SQL generation enables non-experts to interact with databases via natural language. Recent advances rely on large closed-source models like GPT-4 that present challenges in accessibility, privacy, and latency. To address these issues, we focus on developing small, efficient, and open-source text-to-SQL models. We demonstrate the benefits of sampling multiple candidate SQL generations and propose our method, MSc-SQL, to critique them using associated metadata. Our sample critiquing model evaluates multiple outputs simultaneously, achieving state-of-the-art performance compared to other open-source models while remaining competitive with larger models at a much lower cost. Full code can be found at github.com/layer6ai-labs/msc-sql.