Search

Blockchain technology has revolutionized the financial landscape, with cryptocurrencies gaining widespread adoption for their decentralized and transparent nature. As the sentiment expressed on social media platforms can significantly influence cryptocurrency discussions and market movements, sentiment analysis has emerged as a crucial tool for understanding public opinion and predicting market trends. Motivated by the aim to enhance sentiment analysis accuracy in the cryptocurrency domain, this paper investigates fine-tuning techniques on large language models. This paper also investigates the efficacy of supervised fine-tuning and instruction-based fine-tuning on large language models for unseen tasks. Experimental results demonstrate a significant average zero-shot performance gain of 40% after fine-tuning, highlighting the potential of this technique in optimizing pre-trained language model efficiency. Additionally, the impact of instruction tuning on models of varying scales is examined, revealing that larger models benefit from instruction tuning, achieving the highest average accuracy score of 75.16%. In contrast, smaller-scale models may experience reduced generalization due to the complete utilization of model capacity. To gain deeper insight about how instruction works with these language models, this paper presents an experimental investigation into the response of an instruction-based model under different instruction tuning setups. The investigation demonstrates that the model achieves an average accuracy score of 72.38% for short and simple instructions. This performance significantly outperforms its accuracy under long and complex instructions by over 12%, thereby effectively highlighting the profound significance of instruction characteristics in maximizing model performance.

The changes in stromal collagen play a crucial role during the pathogenesis and progression of pancreatic intraepithelial neoplasm (PanIN) to pancreatic ductal adenocarcinoma (PDAC) while misdiagnosis of PanIN is common because of the resemblance to chronic pancreatitis (CP) in its symptoms and subsequent evaluations similarities. To visualize fibrillar collagen in tissues, second harmonic generation microscopy is now utilized as a gold standard in various stromal-based research analyses. However, a technical approach that can perform a quantitative analysis of fibrillar collagen directly on standard slides stained with H&E can (i) discard the need for specialized and costly equipment or labels, (ii) further supplement the conventional histopathological insights and, (iii) potentially be integrated within the framework of standard histopathology workflow. In this study, the whole-core brightfield H&E-stained images of pancreatic tissues were translated computationally into the new collagen images. Subsequently, collagen characteristics of PDAC, PanIN, CP, and normal pancreatic tissues (control) were extracted and compared. The highest alignment (p < 0.01, R2 = 0.2594) was observed in PDAC cores in comparison to the remaining three groups, while the lowest fiber density (p < 0.0001, R2 = 0.3569) was observed in case of normal tissue cores. Moreover, the collagen area and fiber length had shown higher area under curve (0.83 and 0.81, respectively) in discriminating neoplastic and non-neoplastic tissues based on their receiver operating characteristics. The study demonstrated that the computationally generated collagen images can provide a quantitative assessment of collagen remodeling in pancreatic lesions. The cross-modality image synthesis may further lead towards better histopathological and tissue microenvironment insights without the need of specialized imaging equipment or labels.

Modern societies have adopted government-issued fiat currencies many of which exist today mainly in the form of digits in credit and bank accounts. Fiat currencies are controlled by central banks for economic stimulation and stabilization. Boom-and-bust cycles are created. The volatility of the cycle has become increasingly extreme. Social inequality due to the concentration of wealth is prevalent worldwide. As such, restoring sound money, which provides stored value over time, has become a pressing issue. Currently, cryptocurrencies such as Bitcoin are in their infancy and may someday qualify as sound money. Bitcoin today is considered as a digital asset for storing value. But Bitcoin has problems. The first issue of the current Bitcoin network is its high energy consumption consensus mechanism. The second is the cryptographic primitives which are unsafe against post-quantum (PQ) attacks. We aim to propose Green Bitcoin which addresses both issues. To save energy in consensus mechanism, we introduce a post-quantum secure (self-election) verifiable coin-toss function and novel PQ secure proof-of-computation primitives. It is expected to reduce the rate of energy consumption more than 90 percent of the current Bitcoin network. The elliptic curve cryptography will be replaced with PQ-safe versions. The Green Bitcoin protocol will help Bitcoin evolve into a post-quantum secure network., Comment: 16 pages

Magnetic pickup loops or B-dot probes are one of the oldest known sensors of time-varying magnetic fields. The operating principle is based on Faraday's law of electromagnetic induction. However, obtaining accurate measurements of time-varying magnetic fields using these kinds of probes is a challenging task. A B-dot probe and its associated circuit are prone to electrical oscillations. As a result, the measured signal may not faithfully represent the magnetic field sampled by the B-dot probe. In this paper, we have studied the transient response of a B-dot probe and its associated circuit to a time-varying magnetic field. Methods of removing the oscillations pertaining to the detector structure are described. After removing the source of the oscillatory signal, we have shown that the time-integrated induced emf measured by the digitiser is linearly proportional to the magnetic field sampled by the B-dot probe, thus verifying the faithfulness of the measured signal.

We investigate the Beyond Standard Model discovery potential in the framework of the Effective Field Theory (EFT) for the same-sign $WW$ scattering process in purely leptonic $W$ decay modes at the High-Luminosity and High-Energy phases of the Large Hadron Collider (LHC). The goal of this paper is to examine the applicability of the EFT approach, with one dimension-8 operator varied at a time, to describe a hypothetical new physics signal in the $WWWW$ quartic coupling. In the considered process there is no experimental handle on the $WW$ invariant mass, and it has previously been shown that the discovery potential at 14 TeV is rather slim. In this paper we report the results calculated for a 27 TeV machine and compare them with the discovery potential obtained at 14 TeV. We find that while the respective discovery regions shift to lower values of the Wilson coefficients, the overall discovery potential of this procedure does not get significantly larger with a higher beam energy., Comment: 21 pages, published version plus typo corrected in the Acknowledgments