Krisztian Benyo, Ph.D.

📃Scientific paper: A Machine Learning made Catalog of FR-II Radio Galaxies from the FIRST Survey Abstract: We present an independent catalog \(FRIIRGcat\) of 45,241 Fanaroff-Riley Type II \(FR-II\) radio galaxies compiled from the Very Large Array Faint Images of the Radio Sky at Twenty-centimeters \(FIRST\) survey and employed the deep learning method. Among them, optical and/or infrared counterparts are identified for 41,425 FR-IIs. This catalog spans luminosities $2.63\times10^\{22\}\leq L\_\{\rm rad\}\leq6.76\times10^\{29\}\,\{\rm W\}\,\{\rm Hz\}^\{-1\}$ and redshifts up to $z=5.01$. The spectroscopic classification indicates that there are 1431 low-excitation radio galaxies and 260 high-excitation radio galaxies. Among the spectroscopically identified sources, black hole masses are estimated for 4837 FR-IIs, which are in $10^\{7.5\}\lesssim M\_\{\rm BH\}\lesssim 10^\{9.5\}$ $M\_\{\odot\}$. Interestingly, this catalog reveals a couple of giant radio galaxies \(GRGs\), which are already in the existing GRG catalog, confirming the efficiency of this FR-II catalog. Furthermore, 284 new GRGs are unveiled in this new FR-II sample; they have the largest projected sizes ranging from 701 to 1209 kpc and are located at redshifts $0.31\<z\<2.42$. Finally, we explore the distribution of the jet position angle and it shows that the faint Images of the FIRST images are significantly affected by the systematic effect \(the observing beams\). The method presented in this work is expected to be applicable to the radio sky surveys that are currently being conducted because they have finely refined telescope arrays. O... Continued on ES/IODE ➡️ https://etcse.fr/aRX ------- If you find this interesting, feel free to follow, comment and share. We need your help to enhance our visibility, so that our platform continues to serve you.

📃Scientific paper: Large Language Models Offer an Alternative to the Traditional Approach of Topic Modelling Abstract: Topic modelling, as a well-established unsupervised technique, has found extensive use in automatically detecting significant topics within a corpus of documents. However, classic topic modelling approaches (e.g., LDA) have certain drawbacks, such as the lack of semantic understanding and the presence of overlapping topics. In this work, we investigate the untapped potential of large language models (LLMs) as an alternative for uncovering the underlying topics within extensive text corpora. To this end, we introduce a framework that prompts LLMs to generate topics from a given set of documents and establish evaluation protocols to assess the clustering efficacy of LLMs. Our findings indicate that LLMs with appropriate prompts can stand out as a viable alternative, capable of generating relevant topic titles and adhering to human guidelines to refine and merge topics. Through in-depth experiments and evaluation, we summarise the advantages and constraints of employing LLMs in topic extraction. ;Comment: Accepted at LREC-COLING 2024 Continued on ES/IODE ➡️ https://etcse.fr/WmLD ------- If you find this interesting, feel free to follow, comment and share. We need your help to enhance our visibility, so that our platform continues to serve you.

📃Scientific paper: Dynamics for open quantum systems : a discrete quantum level strongly coupled to a continuum with a band structure;Dynamique des systèmes quantiques ouverts : un niveau quantique discret fortement couplé à un continuum avec une structure de bandes Abstract: Following the technological advances of the Industrial Revolution, classical thermodynamics was developed in the 19th century in order to understand the conversion of heat into work in newly designed machines. The works of Boltzmann brought another conceptual revolution with statistical mechanics. He demonstrated the microscopical origin of the laws of thermodynamics which actually only describe the macroscopic behaviour of systems in which local thermalization is faster than all other timescales. However, following the growing interest for nanotechnologies, it is now possible to manipulate microscopic systems in which thermalization is slower than the timescales for electron flow. A major technological advance in this field stems from the use of quantum dots, nanoscale devices which confine electrons on such small scales that they spread on discrete energy levels. It is then essential to take into account quantum effects for the study of this type of systems, that is to say to design theoretical tools combining thermodynamics and quantum mechanics.Problems of quantum thermodynamics are often tackled in the framework of the theory of open quantum systems. The general idea of this formalism is to study the dynamics of a “small” quantum system when it is coupled to another much bigger representing the environment. One can then show that the time evolution of the small system can be described by a master equation in the limit where it is weakly coupled to the environmen... Continued on ES/IODE ➡️ https://etcse.fr/n8A ------- If you find this interesting, feel free to follow, comment and share. We need your help to enhance our visibility, so that our platform continues to serve you.

📃Scientific paper: CaMML: Context-Aware Multimodal Learner for Large Models Abstract: In this work, we introduce Context-Aware MultiModal Learner (CaMML), for tuning large multimodal models (LMMs). CaMML, a lightweight module, is crafted to seamlessly integrate multimodal contextual samples into large models, thereby empowering the model to derive knowledge from analogous, domain-specific, up-to-date information and make grounded inferences. Importantly, CaMML is highly scalable and can efficiently handle lengthy multimodal context examples owing to its hierarchical design. Based on CaMML, we have developed two multimodal models, CaMML-7B and CaMML-13B, that have shown exceptional performance across an array of benchmark datasets for multimodal tasks. Remarkably, CaMML-13B achieves the state-of-the-art performance on over ten widely recognized multimodal benchmark datasets, surpassing LLaVA-1.5 (13B) with a noticeable margin, without integration of any external resources. Moreover, we have conducted extensive ablative studies to inspect the inner workings of CaMML and performed qualitative analyses to showcase its effectiveness in handling real-world challenging cases. Code and models are available at: https://lnkd.in/eE5k3xrf. ;Comment: Preprint Continued on ES/IODE ➡️ https://etcse.fr/HL2s ------- If you find this interesting, feel free to follow, comment and share. We need your help to enhance our visibility, so that our platform continues to serve you.

📃Scientific paper: Experimenting with Large Language Models and vector embeddings in NASA SciX Abstract: Open-source Large Language Models enable projects such as NASA SciX (i.e., NASA ADS) to think out of the box and try alternative approaches for information retrieval and data augmentation, while respecting data copyright and users' privacy. However, when large language models are directly prompted with questions without any context, they are prone to hallucination. At NASA SciX we have developed an experiment where we created semantic vectors for our large collection of abstracts and full-text content, and we designed a prompt system to ask questions using contextual chunks from our system. Based on a non-systematic human evaluation, the experiment shows a lower degree of hallucination and better responses when using Retrieval Augmented Generation. Further exploration is required to design new features and data augmentation processes at NASA SciX that leverages this technology while respecting the high level of trust and quality that the project holds. ;Comment: To appear in the proceedings of the 33th annual international Astronomical Data Analysis Software & Systems (ADASS XXXIII) Continued on ES/IODE ➡️ https://etcse.fr/MY2st ------- If you find this interesting, feel free to follow, comment and share. We need your help to enhance our visibility, so that our platform continues to serve you.

📃Scientific paper: An application of neighbourhoods in digraphs to the classification of binary dynamics Abstract: A binary state on a graph means an assignment of binary values to its vertices. A time-dependent sequence of binary states is referred to as binary dynamics. We describe a method for the classification of binary dynamics of digraphs, using particular choices of closed neighbourhoods. Our motivation and application comes from neuroscience, where a directed graph is an abstraction of neurons and their connections, and where the simplification of large amounts of data is key to any computation. We present a topological/graph theoretic method for extracting information out of binary dynamics on a graph, based on a selection of a relatively small number of vertices and their neighbourhoods. We consider existing and introduce new real-valued functions on closed neighbourhoods, comparing them by their ability to accurately classify different binary dynamics. We describe a classification algorithm that uses two parameters and sets up a machine learning pipeline. We demonstrate the effectiveness of the method on simulated activity on a digital reconstruction of cortical tissue of a rat, and on a nonbiological random graph with similar density. Continued on ES/IODE ➡️ https://etcse.fr/Awc ------- If you find this interesting, feel free to follow, comment and share. We need your help to enhance our visibility, so that our platform continues to serve you.

The paper builds on Einstein’s work from 1928 on “teleparallel gravity”, an alternative interpretation of gravity that considers both the curvature and torsion of spacetime... https://lnkd.in/eY4ifTVh

📃Scientific paper: Audiovisual Highlight Detection in Videos Abstract: In this paper, we test the hypothesis that interesting events in unstructured videos are inherently audiovisual. We combine deep image representations for object recognition and scene understanding with representations from an audiovisual affect recognition model. To this set, we include content agnostic audio-visual synchrony representations and mel-frequency cepstral coefficients to capture other intrinsic properties of audio. These features are used in a modular supervised model. We present results from two experiments: efficacy study of single features on the task, and an ablation study where we leave one feature out at a time. For the video summarization task, our results indicate that the visual features carry most information, and including audiovisual features improves over visual-only information. To better study the task of highlight detection, we run a pilot experiment with highlights annotations for a small subset of video clips and fine-tune our best model on it. Results indicate that we can transfer knowledge from the video summarization task to a model trained specifically for the task of highlight detection. ;Comment: 5 pages, 2 figures, conference paper Continued on ES/IODE ➡️ https://etcse.fr/nFoAt ------- If you find this interesting, feel free to follow, comment and share. We need your help to enhance our visibility, so that our platform continues to serve you.

📃Scientific paper: Strings from the Library of Babel: Random Sampling as a Strong Baseline for Prompt Optimisation Abstract: Recent prompt optimisation approaches use the generative nature of language models to produce prompts -- even rivaling the performance of human-curated prompts. In this paper, we demonstrate that randomly sampling tokens from the model vocabulary as ``separators'' can be as effective as language models for prompt-style text classification. Our experiments show that random separators are competitive baselines, having less than a 1% difference compared to previous self-optimisation methods and showing a 12% average relative improvement over strong human baselines across nine text classification tasks and eight language models. We further analyse this phenomenon in detail using three different random generation strategies, establishing that the language space is rich with potentially good separators, with a greater than 40% average chance that a randomly drawn separator performs better than human-curated separators. These observations challenge the common assumption that an effective prompt should be human readable or task relevant and establish a strong baseline for prompt optimisation research. ;Comment: Accepted to NAACL 2024. The code is publicly available at https://lnkd.in/dvgZn2SX Continued on ES/IODE ➡️ https://etcse.fr/Cm ------- If you find this interesting, feel free to follow, comment and share. We need your help to enhance our visibility, so that our platform continues to serve you.