Despite their improved capabilities in generation and reasoning, adapting large language models (LLMs) to the biomedical domain remains challenging due to their immense size and corporate privacy. In this work, we propose MedAdapter, a unified post-hoc adapter for test-time adaptation of LLMs towards biomedical applications. Instead of fine-tuning the entire LLM, MedAdapter effectively adapts the original model by fine-tuning only a small BERT-sized adapter to rank candidate solutions generated by LLMs. Experiments demonstrate that MedAdapter effectively adapts both white-box and black-box LLMs in biomedical reasoning, achieving average performance improvements of 25.48% and 11.31%, respectively, without requiring extensive computational resources or sharing data with third parties. MedAdapter also yields superior performance when combined with train-time adaptation, highlighting a flexible and complementary solution to existing adaptation methods. Faced with the challenges of balancing model performance, computational resources, and data privacy, MedAdapter provides an efficient, privacy-preserving, cost-effective, and transparent solution for adapting LLMs to the biomedical domain.

Numerous sign language datasets exist, yet they typically cover only a limited selection of the thousands of signs used globally. Moreover, creating diverse sign language datasets is an expensive and challenging task due to the costs associated with gathering a varied group of signers. Motivated by these challenges, we aimed to develop a solution that addresses these limitations. In this context, we focused on textually describing body movements from skeleton keypoint sequences, leading to the creation of a new dataset. We structured this dataset around AUTSL, a comprehensive isolated Turkish sign language dataset. We also developed a baseline model, SkelCap, which can generate textual descriptions of body movements. This model processes the skeleton keypoints data as a vector, applies a fully connected layer for embedding, and utilizes a transformer neural network for sequence-to-sequence modeling. We conducted extensive evaluations of our model, including signer-agnostic and sign-agnostic assessments. The model achieved promising results, with a ROUGE-L score of 0.98 and a BLEU-4 score of 0.94 in the signer-agnostic evaluation. The dataset we have prepared, namely the AUTSL-SkelCap, will be made publicly available soon.

Large bibliometric databases, such as Web of Science, Scopus, and OpenAlex, facilitate bibliometric analyses, but are performative, affecting the visibility of scientific outputs and the impact measurement of participating entities. Recently, these databases have taken up the UN's Sustainable Development Goals (SDGs) in their respective classifications, which have been criticised for their diverging nature. This work proposes using the feature of large language models (LLMs) to learn about the "data bias" injected by diverse SDG classifications into bibliometric data by exploring five SDGs. We build a LLM that is fine-tuned in parallel by the diverse SDG classifications inscribed into the databases' SDG classifications. Our results show high sensitivity in model architecture, classified publications, fine-tuning process, and natural language generation. The wide arbitrariness at different levels raises concerns about using LLM in research practice.

Low-rank adaptation~(LoRA) has recently gained much interest in fine-tuning foundation models. It effectively reduces the number of trainable parameters by incorporating low-rank matrices $A$ and $B$ to represent the weight change, i.e., $\Delta W=BA$. Despite LoRA's progress, it faces storage challenges when handling extensive customization adaptations or larger base models. In this work, we aim to further compress trainable parameters by enjoying the powerful expressiveness of the Fourier transform. Specifically, we introduce FourierFT, which treats $\Delta W$ as a matrix in the spatial domain and learns only a small fraction of its spectral coefficients. With the trained spectral coefficients, we implement the inverse discrete Fourier transform to recover $\Delta W$. Empirically, our FourierFT method shows comparable or better performance with fewer parameters than LoRA on various tasks, including natural language understanding, natural language generation, instruction tuning, and image classification. For example, when performing instruction tuning on the LLaMA2-7B model, FourierFT surpasses LoRA with only 0.064M trainable parameters, compared to LoRA's 33.5M. Our code is released at \url{https://github.com/Chaos96/fourierft}.

Incremental learning is the ability of systems to acquire knowledge over time, enabling their adaptation and generalization to novel tasks. It is a critical ability for intelligent, real-world systems, especially when data changes frequently or is limited. This review provides a comprehensive analysis of incremental learning in Large Language Models. It synthesizes the state-of-the-art incremental learning paradigms, including continual learning, meta-learning, parameter-efficient learning, and mixture-of-experts learning. We demonstrate their utility for incremental learning by describing specific achievements from these related topics and their critical factors. An important finding is that many of these approaches do not update the core model, and none of them update incrementally in real-time. The paper highlights current problems and challenges for future research in the field. By consolidating the latest relevant research developments, this review offers a comprehensive understanding of incremental learning and its implications for designing and developing LLM-based learning systems.

In real-world scenarios, objects often require repositioning and reorientation before they can be grasped, a process known as pre-grasp manipulation. Learning universal dexterous functional pre-grasp manipulation requires precise control over the relative position, orientation, and contact between the hand and object while generalizing to diverse dynamic scenarios with varying objects and goal poses. To address this challenge, we propose a teacher-student learning approach that utilizes a novel mutual reward, incentivizing agents to optimize three key criteria jointly. Additionally, we introduce a pipeline that employs a mixture-of-experts strategy to learn diverse manipulation policies, followed by a diffusion policy to capture complex action distributions from these experts. Our method achieves a success rate of 72.6\% across more than 30 object categories by leveraging extrinsic dexterity and adjusting from feedback.

Learning an ordering of items based on noisy pairwise comparisons is useful when item-specific labels are difficult to assign, for example, when annotators have to make subjective assessments. Algorithms have been proposed for actively sampling comparisons of items to minimize the number of annotations necessary for learning an accurate ordering. However, many ignore shared structure between items, treating them as unrelated, limiting sample efficiency and precluding generalization to new items. In this work, we study active learning with pairwise preference feedback for ordering items with contextual attributes, both in- and out-of-sample. We give an upper bound on the expected ordering error incurred by active learning strategies under a logistic preference model, in terms of the aleatoric and epistemic uncertainty in comparisons, and propose two algorithms designed to greedily minimize this bound. We evaluate these algorithms in two realistic image ordering tasks, including one with comparisons made by human annotators, and demonstrate superior sample efficiency compared to non-contextual ranking approaches and active preference learning baselines.

Nowadays, there is a rapid development of information technology, which entails the need to constantly improve and expand the capabilities of interactive artificial intelligence systems This monograph combines several current topics related to the field of information technology One of the key topics is the methodology for enhancing the capabilities of conversational systems, with a focus on ChatGPT, which represents the latest advance in the field of artificial intelligence The monograph also discusses text generation systems based on ontological representations, which open up wide opportunities for creating high-quality content A special place in the work is given to an automated computer system for diagnosing the competitiveness of specialists in the field of information technology This helps to effectively assess the professionalism of specialists and determine the need for advanced training Theoretical aspects of correct color rendering and informatization of educational and research work of graduate students are important in ensuring the quality of education and scientific research And finally, the use of technology for creating 3D models has become an integral part of the modern information environment, which makes it possible to bring the most daring ideas and projects to life Research and development in these areas contribute to the improvement of information technologies, finding application in various fields of activity The purpose of our monograph is to conduct analysis and research in these areas in order to promote the development of information technologies and increase their efficiency The monograph was compiled based on the results of the XVI international scientific and practical conference "Information technologies and automation -- 2023", which took place in October 2023 at Odessa National University of Technology

While Large Language Models (LLMs) have demonstrated significant promise as agents in interactive tasks, their substantial computational requirements and restricted number of calls constrain their practical utility, especially in long-horizon interactive tasks such as decision-making or in scenarios involving continuous ongoing tasks. To address these constraints, we propose a method for transferring the performance of an LLM with billions of parameters to a much smaller language model (770M parameters). Our approach involves constructing a hierarchical agent comprising a planning module, which learns through Knowledge Distillation from an LLM to generate sub-goals, and an execution module, which learns to accomplish these sub-goals using elementary actions. Subsequently, we utilize this annotated data to fine-tune both the planning and execution modules. Importantly, neither module relies on real-time access to an LLM during inference, significantly reducing the overall cost associated with LLM interactions to a fixed cost. In ScienceWorld, a challenging and multi-task interactive text environment, our method surpasses standard imitation learning based solely on elementary actions by 16.7% (absolute). Our analysis highlights the efficiency of our approach compared to other LLM-based methods. Recently, Large Language Models (LLMs) have found applications in various fields, including multi-task learning, decision making, answering questions, summarizing documents, translating languages, completing sentences, and serving as search assistants. The promising advantage of LLMs is attributed to their training on extensive text datasets, resulting in impressive capabilities. This prior knowledge can be leveraged for action planning to solve tasks in robotics and reinforcement learning (Huang et al., 2022b; Brohan et al., 2023; Liang et al., 2023). However, the extreme size of LLMs makes them computationally unaffordable for many applications. Consequently, there is an increasing demand to find approaches that are less computationally intensive while still capitalizing on the knowledge embedded in LLMs. One prevalent technique is the use of Knowledge Distillation (KD) (Buciluǎ et al., 2006; Hinton et al., 2015), wherein a smaller model is trained with guidance from a larger model. Through this approach, we can leverage the knowledge in an LLM to train a more compact model with a reduced number of parameters. First, focus on the substance. Figure 1: Example of annotating an expert trajectory with sub-goals for a particular variation of task 1-4 We employ Knowledge Distillation from an LLM to train (change-the-state-of-matter-of).

Generative Artificial Intelligence (AI) is integrated into everyday technology, including news, education, and social media. AI has further pervaded private conversations as conversational partners, auto-completion, and response suggestions. As social media becomes young people's main method of peer support exchange, we need to understand when and how AI can facilitate and assist in such exchanges in a beneficial, safe, and socially appropriate way. We asked 622 young people to complete an online survey and evaluate blinded human- and AI-generated responses to help-seeking messages. We found that participants preferred the AI-generated response to situations about relationships, self-expression, and physical health. However, when addressing a sensitive topic, like suicidal thoughts, young people preferred the human response. We also discuss the role of training in online peer support exchange and its implications for supporting young people's well-being. Disclaimer: This paper includes sensitive topics, including suicide ideation. Reader discretion is advised.