Automating network processes without human intervention is crucial for the complex Sixth Generation (6G) environment. Thus, 6G networks must advance beyond basic automation, relying on Artificial Intelligence (AI) and Machine Learning (ML) for self-optimizing and autonomous operation. This requires zero-touch management and orchestration, the integration of Network Intelligence (NI) into the network architecture, and the efficient lifecycle management of intelligent functions. Despite its potential, integrating NI poses challenges in model development and application. To tackle those issues, this paper presents a novel methodology to manage the complete lifecycle of Reinforcement Learning (RL) applications in networking, thereby enhancing existing Machine Learning Operations (MLOps) frameworks to accommodate RL-specific tasks. We focus on scaling computing resources in service-based architectures, modeling the problem as a Markov Decision Process (MDP). Two RL algorithms, guided by distinct Reward Functions (RFns), are proposed to autonomously determine the number of service replicas in dynamic environments. Our proposed methodology is anchored on a dual approach: firstly, it evaluates the training performance of these algorithms under varying RFns, and secondly, it validates their performance after being trained to discern the practical applicability in real-world settings. We show that, despite significant progress, the development stage of RL techniques for networking applications, particularly in scaling scenarios, still leaves room for significant improvements. This study underscores the importance of ongoing research and development to enhance the practicality and resilience of RL techniques in real-world networking environments.

The cleavage of C-S bonds represents a crucial step in fossil fuel refinement to remove organosulfur impurities. Efforts are required to identify alternatives that can replace the energy-intensive hydrodesulfurization process currently in use. In this context, we have developed a series of bis-thiolato-ligated CrIII complexes supported by the L2- ligand (L2- = 2,2'-bipyridine-6,6'-diyl(bis(1,1-diphenylethanethiolate), one of them displaying desulfurization of one thiolate of the ligand under reducing and acidic conditions at ambient temperature and atmospheric pressure. While only 5-coordinated complexes were previously isolated by reaction of L2- with 3d metal MIII ions, both 5- and 6-coordinated mononuclear complexes have been obtained in the case of CrIII, viz., [CrIIILCl], [CrIIILCl2]-, and [CrIIILCl(CH3CN)]. The investigation of the reactivity of [CrIIILCl(CH3CN)] under reducing conditions led to a dinuclear [CrIII2L2(μ-Cl)(μ-OH)] compound and, in the presence of protons, to the mononuclear CrIII complex [CrIII(LN2S)2]+, where LN2S- is the partially desulfurized form of L2-. A desulfurization mechanism has been proposed involving the release of H2S, as evidenced experimentally.

Large language models (LLMs) have revolutionized Natural Language Processing (NLP), but their size creates computational bottlenecks. We introduce a novel approach to create accurate, sparse foundational versions of performant LLMs that achieve full accuracy recovery for fine-tuning tasks at up to 70% sparsity. We achieve this for the LLaMA-2 7B model by combining the SparseGPT one-shot pruning method and sparse pretraining of those models on a subset of the SlimPajama dataset mixed with a Python subset of The Stack dataset. We exhibit training acceleration due to sparsity on Cerebras CS-3 chips that closely matches theoretical scaling. In addition, we establish inference acceleration of up to 3x on CPUs by utilizing Neural Magic's DeepSparse engine and 1.7x on GPUs through Neural Magic's nm-vllm engine. The above gains are realized via sparsity alone, thus enabling further gains through additional use of quantization. Specifically, we show a total speedup on CPUs for sparse-quantized LLaMA models of up to 8.6x. We demonstrate these results across diverse, challenging tasks, including chat, instruction following, code generation, arithmetic reasoning, and summarization to prove their generality. This work paves the way for rapidly creating smaller and faster LLMs without sacrificing accuracy.

This paper analyzes the influence of foliar fertilizer based on humus extract on some of the elements central to the quality of the Polka raspberry variety in the area of the city of Bihać. The research was conducted in 2015 according to the control and treatment system. A foliar fertilizer based on humus extract was used for the treatment. A total of 12 quantitative and qualitative properties were analyzed: content of total sugars, reducing sugars, invert sugars, sucrose, water content, dry matter, total acidity, vitamin C, total phenols, total flavonoids, antioxidant capacity, and fruit mass. After the analyses were completed, it can be concluded that fruits of raspberry plants treated with humus extract have an almost identical value of total acidity (1.65%) as found in control. However, treatment with a humus extract slightly lowered the content of total phenols and vitamin C. Based on the results obtained, it was shown that the foliar fertilizer based on the humus extract did not have a positive effect on certain chemical and antioxidant properties of the raspberry fruit.

Honey is a natural product that is an excellent source of energy containing mainly carbohydrates and water, as well as small amounts of organic acids, vitamins, minerals, flavonoids, and enzymes. Due to the presence of bioactive compounds, it has been shown that honey is beneficial for many diseases, e.g. gastrointestinal diseases, skin diseases, cancer, heart diseases, and neurological degeneration. The study of the physical and chemical properties of honey and the content of bioactive compounds has been increasingly applied in order to determine the quality of honey samples. The aim of this study is to investigate physicochemical properties as well as the total phenol content and antioxidant activity of seven multifloral honey samples from the Herzegovina region. Physicochemical parameters determined in the honey samples (moisture, acidity, electrical conductivity, reducing sugars, sucrose, and insoluble matter) were within the quality standard limits of the Regulation on methods for control of honey and other bee products. Total phenolic content was determined using the Folin-Ciocalteu method and it ranged from 46.98 ± 6.36 to 152.94 ± 4.95 mg GAE/100 g of honey. To determine the antioxidant activity of the honey samples, two methods, FRAP and ABTS, were used. The total phenolic content of honey correlated positively with its antioxidant activity.

Klebsiella pneumoniae, a member of the Enterobacteriaceae family, demonstrates an increasing trend of resistance to carbapenems and is a common cause of both hospital- and community-acquired infections. The current study provides insights into the genetic characterization of carbapenem-resistant K. pneumoniae (CRKP) isolates circulating during 2022 in a Sarajevo tertiary hospital. Among the 87 CRKP strains analyzed, real-time polymerase chain reaction (rtPCR) results showed that 85 (97.7%) tested positive for the carbapenem resistance gene. The oxacillinase-48 (OXA-48) gene was detected in 83 (95.4%) isolates, while the K. pneumoniae carbapenemase (KPC) and the New Delhi metallo-beta-lactamase (NDM) genes were detected in one isolate each. No Verona integron-encoded-metallo-beta-lactamase (VIM) or imipenemase-metallo-beta-lactamase 1 (IMP-1) genes were found in any of the tested isolates. The multilocus sequence typing (MLST) analysis of sequence types (STs) revealed that ST101, an emerging high-risk clone exhibiting extensive drug resistance, was the most prevalent, whereas ST307 was detected in only one isolate. Phylogenetic analysis of the ten CRKP isolates indicated the presence of three clusters that could constitute an outbreak. A comparison of the results of the utilized phenotypic test (the combined-disk test [CDT]) and rtPCR showed high concordance, suggesting that the phenotypic assay may be useful for the early detection of resistance mechanisms as part of routine susceptibility testing. With the increased affordability of next-generation sequencing (NGS), its application in hospital settings has proven highly beneficial, aiding in the implementation of infection control and prevention measures. Given the significant resistance demonstrated by the CRKP isolates to most tested antibiotics, it is imperative to establish effective methods to restrict the spread of these isolates, as well as to carefully monitor the use of carbapenems in clinical practice.

Chlorogenic acid, an ester of caffeic acid with quinic acid, also known as 5- O-caffeoylquinic acid (5-CQA), is a ubiquitous plant constituent that is an important intermediate in lignin biosynthesis. In some cases, it occurs at surprisingly high levels in the leaves and fruits of certain higher plants, such as coffee beans. Due to its catechol moiety and an extended side chain conjugation, it easily forms a resonance-stabilised phenoxy radical, accounting for its powerful antioxidant potential. The objective of this work was to determine if the esterification and methylation of 5- CQA would enhance its antioxidant activity. Two 5-CQA derivatives were prepared for this study. Chlorogenic acid was esterified with methanol over Amberlite IR120-H to obtain methyl chlorogenate, while methyl 3',4´-dimethyl chlorogenate was prepared from 5-CQA by treatment with diazomethane. Spectroscopic methods confirmed the structure of these derivatives. Their antioxidant properties were tested to establish a relationship between structure and antioxidant activity. Antioxidant activity results were generated for 5-CQA and its ester analogues using eight different methods. Depending on the method applied, results were expressed as IC50/MCE50 values or as equivalents of the applied standard (ascorbic acid and Trolox). In most of these tests, 5-CQA showed the highest antioxidant activity compared to its derivatives. Nevertheless, due to their hydrophobic characteristics, their ester analogues remain promising antioxidant candidates in emulsifying systems.

Introduction: Meningiomas are the most common benign tumor of the central nervous system, accounting for 53.3% and 37.6% of all central nervous system tumors (1). The World Health Organization (WHO) Grade I meningiomas account for 80.5% of all meningiomas and are considered benign meningiomas; the WHO Grade II meningiomas account for 17.7% of all meningiomas and exhibit more aggressive behavior. Methods: In the period 2015-2022, a retrospective single-center study at the clinic of neurosurgery at the Clinical Center University of Sarajevo was conducted, which included patients with a pathohistological finding of WHO Grade I or II meningioma. Depending on the pathohistological grade of the tumor, patients were divided into two groups: Grade I and Grade II patients. Patients were examined clinically and radiologically. Clinical data collected included in the study: Gender, age, number of symptoms before surgery, whether patients were symptomatic or asymptomatic, pre-operative Eastern Cooperative Oncology Group,and Karnopsky performance scale. Pre-operative contrast magnetic resonance imaging of the head measured tumor volume, temporal muscle thickness (TMT), sagittal midline shift, and surrounding cerebral edema. Results: A total of 80 patients were enrolled in the study, 68 with WHO Grade I and 12 with WHO Grade II meningiomas. We found that patients with Grade I meningioma were younger and that the mean thickness of the temporal muscle was statistically thicker than in patients with Grade II. Increasing TMT was significantly and positively associated with Grade I tumors and negatively associated with Grade II tumors (p = 0.032). Conclusion: This study demonstrates that TMT can serve as a radiologic pre-operative indicator of meningioma grade and provide valuable guidance to neurosurgeons in surgical planning. Further studies are needed to validate these results.

This study aimed to investigate the impact of an 8-week aerobic dance intervention on postural balance in children. Forty-one children, aged 9 to 11, were randomly assigned to either an aerobic dance group (ADG) or a control group (CG) from a primary school. Postural balance was assessed using center of pressure (CoP) excursions before and after the 8-week intervention period. Evaluations were conducted on both firm and foam surfaces in bipedal and unipedal stances under open-eyes (OE) and closed-eyes (CE) conditions, as well as on both medial–lateral (ML) and anterior–posterior (AP) surfaces in a bipedal stance under OE conditions. The ADG exhibited significantly decreased CoPVm values during firm bipedal CE, unipedal OE, foam bipedal OE and CE, and foam unipedal OE (p < 0.005). This study suggests that aerobic dance intervention improved postural balance in children, showcasing adaptability and improved stability under various conditions.