Natural language can offer a concise and human-interpretable means of specifying reinforcement learning (RL) tasks. The ability to extract rewards from a language instruction can enable the development of robotic systems that can learn from human guidance; however, it remains a challenging problem, especially in visual environments. Existing approaches that employ large, pretrained language models either rely on non‑visual environment representations, require prohibitively large amounts of feedback, or generate noisy, ill‑shaped reward functions. In this paper, we propose a novel method, GoalLadder, that leverages vision-language models (VLMs) to train RL agents from a single language instruction in visual environments. GoalLadder works by incrementally discovering states that bring the agent closer to completing a task specified in natural language. To do so, it queries a VLM to identify states that represent an improvement in agent's task progress and to rank them using pairwise comparisons. Unlike prior work, GoalLadder does not trust VLM's feedback completely; instead, it uses it to rank potential goal states using an ELO-based rating system, thus reducing the detrimental effects of noisy VLM feedback. Over the course of training, the agent is tasked with minimising the distance to the top-ranked goal in a learned embedding space, which is trained on unlabelled visual data. This key feature allows us to bypass the need for abundant and accurate feedback typically required to train a well-shaped reward function. We demonstrate that GoalLadder outperforms existing related methods on classic control and robotic manipulation environments with the average final success rate of $\sim$95\% compared to only $\sim$45\% of the best competitor.

MOTIVATION: The transcriptomic data are being frequently used in the research of biomarker genes of different diseases and biological states. The most common tasks there are the data harmonization and treatment outcome prediction. Both of them can be addressed via the style transfer approach. Either technical factors or any biological details about the samples which we would like to control (gender, biological state, treatment, etc.) can be used as style components. RESULTS: The proposed style transfer solution is based on Conditional Variational Autoencoders, Y-Autoencoders and adversarial feature decomposition. To quantitatively measure the quality of the style transfer, neural network classifiers which predict the style and semantics after training on real expression were used. Comparison with several existing style-transfer based approaches shows that proposed model has the highest style prediction accuracy on all considered datasets while having comparable or the best semantics prediction accuracy. AVAILABILITY AND IMPLEMENTATION: https://github.com/NRshka/stvae-source. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Most of the current state of the art motion planning software libraries are not easily interchangeable, because core concepts are represented with different data structures, application programming interfaces (API) are not compatible, or algorithms are encapsulated in modules organized in mutually exclusive abstraction hierarchies. These problems limit the possibility to reuse different libraries in the same application interchangeably and to compare their quality attributes (performance, completeness, etc.). An approach to overcome these shortcomings is refactoring, a technique that aims to restructure a set of existing software libraries without affecting their external behavior in order to harmonize their architecture, data structures, and APIs. This paper presents a component-based refactoring approach that has allowed the transition from motion planning libraries, taking the object-oriented framework CoPP as basis, to a component-based system. In particular we describe a four-step application of well-known architecture refactoring patterns that redistributes the responsibilities among the classes, harmonizes the common data structures and reduces the coupling degree. The obtained system represents a composition of reusable components that are easy to customize and offer different algorithms to resolve the same problem. In this way the user could quickly compile a new motion planning application by simply choosing which algorithm to use for each functionality.