Przemyslaw Kazienko

Wei Jie Yeo, Teddy Ferdinan, Przemyslaw Kazienko et al.

Large language models (LLMs) often necessitate extensive labeled datasets and training compute to achieve impressive performance across downstream tasks. This paper explores a self-training paradigm, where the LLM autonomously curates its own labels and selectively trains on unknown data samples identified through a reference-free consistency method. Empirical evaluations demonstrate significant improvements in reducing hallucination in generation across multiple subjects. Furthermore, the selective training framework mitigates catastrophic forgetting in out-of-distribution benchmarks, addressing a critical limitation in training LLMs. Our findings suggest that such an approach can substantially reduce the dependency on large labeled datasets, paving the way for more scalable and cost-effective language model training.

Bo Peng, Eric Alcaide, Quentin Anthony et al.

Transformers have revolutionized almost all natural language processing (NLP) tasks but suffer from memory and computational complexity that scales quadratically with sequence length. In contrast, recurrent neural networks (RNNs) exhibit linear scaling in memory and computational requirements but struggle to match the same performance as Transformers due to limitations in parallelization and scalability. We propose a novel model architecture, Receptance Weighted Key Value (RWKV), that combines the efficient parallelizable training of transformers with the efficient inference of RNNs. Our approach leverages a linear attention mechanism and allows us to formulate the model as either a Transformer or an RNN, thus parallelizing computations during training and maintains constant computational and memory complexity during inference. We scale our models as large as 14 billion parameters, by far the largest dense RNN ever trained, and find RWKV performs on par with similarly sized Transformers, suggesting future work can leverage this architecture to create more efficient models. This work presents a significant step towards reconciling trade-offs between computational efficiency and model performance in sequence processing tasks.

Katarzyna Pasierb, Tomasz Kajdanowicz, Przemyslaw Kazienko

The goal of the paper is to present different approaches to privacy-preserving data sharing and publishing in the context of e-health care systems. In particular, the literature review on technical issues in privacy assurance and current real-life high complexity implementation of medical system that assumes proper data sharing mechanisms are presented in the paper.

Tomasz Kajdanowicz, Przemyslaw Kazienko, Piotr Doskocz

A new method of feature extraction in the social network for within-network classification is proposed in the paper. The method provides new features calculated by combination of both: network structure information and class labels assigned to nodes. The influence of various features on classification performance has also been studied. The experiments on real-world data have shown that features created owing to the proposed method can lead to significant improvement of classification accuracy.

Przemyslaw Kazienko, Katarzyna Musial, Tomasz Kajdanowicz

All online sharing systems gather data that reflects users' collective behaviour and their shared activities. This data can be used to extract different kinds of relationships, which can be grouped into layers, and which are basic components of the multidimensional social network proposed in the paper. The layers are created on the basis of two types of relations between humans, i.e. direct and object-based ones which respectively correspond to either social or semantic links between individuals. For better understanding of the complexity of the social network structure, layers and their profiles were identified and studied on two, spanned in time, snapshots of the Flickr population. Additionally, for each layer, a separate strength measure was proposed. The experiments on the Flickr photo sharing system revealed that the relationships between users result either from semantic links between objects they operate on or from social connections of these users. Moreover, the density of the social network increases in time. The second part of the study is devoted to building a social recommender system that supports the creation of new relations between users in a multimedia sharing system. Its main goal is to generate personalized suggestions that are continuously adapted to users' needs depending on the personal weights assigned to each layer in the multidimensional social network. The conducted experiments confirmed the usefulness of the proposed model.