We thank all the reviewers for their valuable comments. We would like to clarify that, 'When the model was trained without the mel-spectrogram loss, the training process We also think that applying the L1/L2 loss gives no disadvantage in one-to-one mapping as our work. We will clarify the details of the experiments in Section 3. Table 1: Mean Opinion Scores. All models were trained up to 500k steps. MOS evaluation results are shown in [Table 1].

LLM pruning has emerged as a promising technology for compressing LLMs, enabling their deployment on resource-limited devices. However, current methodologies typically require access to public calibration samples, which can be challenging to obtain in privacy-sensitive domains. To address this issue, we introduce FedPrLLM, a comprehensive federated pruning framework designed for the privacy-preserving compression of LLMs. In FedPrLLM, each client only needs to calculate a pruning mask matrix based on its local calibration data and share it with the server to prune the global model. This approach allows for collaborative pruning of the global model with the knowledge of each client while maintaining local data privacy. Additionally, we conduct extensive experiments to explore various possibilities within the FedPrLLM framework, including different comparison groups, pruning strategies, and the decision to scale weights. Our extensive evaluation reveals that one-shot pruning with layer comparison and no weight scaling is the optimal choice within the FedPrLLM framework. We hope our work will help guide future efforts in pruning LLMs in privacy-sensitive fields. Our code is available at https://github.com/Pengxin-Guo/FedPrLLM.

As their size increases, Large Languages Models (LLMs) are natural candidates for network pruning methods: approaches that drop a subset of network weights while striving to preserve performance. Existing methods, however, require either retraining, which is rarely affordable for billion-scale LLMs, or solving a weight reconstruction problem reliant on second-order information, which may also be computationally expensive. In this paper, we introduce a novel, straightforward yet effective pruning method, termed Wanda (Pruning by Weights and activations), designed to induce sparsity in pretrained LLMs. Motivated by the recent observation of emergent large magnitude features in LLMs, our approach prunes weights with the smallest magnitudes multiplied by the corresponding input activations, on a per-output basis. Notably, Wanda requires no retraining or weight update, and the pruned LLM can be used as is. We conduct a thorough evaluation of our method Wanda on LLaMA and LLaMA-2 across various language benchmarks. Wanda significantly outperforms the established baseline of magnitude pruning and performs competitively against recent method involving intensive weight update. Code is available at https://github.com/locuslab/wanda.

Due to the precautionary measures during the COVID-19 pandemic many universities offered unproctored take-home exams. We propose methods to detect potential collusion between students and apply our approach on event log data from take-home exams during the pandemic. We find groups of students with suspiciously similar exams. In addition, we compare our findings to a proctored control group. By this, we establish a rule of thumb for evaluating which cases are "outstandingly similar", i.e., suspicious cases.

This article introduces and implements the framework of propensity score method from Dehejia and Wahba (1999) "Causal Effects in Non-Experimental Studies: Reevaluating the Evaluation of Training Programs," Journal of the American Statistical Association, Vol. I will briefly go over the theories and then walk through how I implemented the stratification matching step by step. The full Python code is provided at the end of the article. The intuition of propensity score method is: instead of conditioning on the full vector of covariates Xᵢ, which can get difficult when there are many pre-treatment variables and when the treatment and comparison groups are very different, we try to condition on the propensity score estimated with Xᵢ. Propensity score matching works in the same way as covariate matching except that we match on the score instead of the covariates directly.

Desperate to solve the deadly conundrum of COVID-19, the world is clamoring for fast answers and solutions from a research system not built for haste. The ironic, and perhaps tragic, result: Scientific shortcuts have slowed understanding of the disease and delayed the ability to find out which drugs help, hurt or have no effect at all. As deaths from the coronavirus relentlessly mounted into the hundreds of thousands, tens of thousands of doctors and patients rushed to use drugs before they could be proved safe or effective. "People had an epidemic in front of them and were not prepared to wait," said Dr. Derek Angus, critical care chief at the University of Pittsburgh Medical Center. "We made traditional clinical research look so slow and cumbersome."

The UAE is an early adopter of advanced technology, and artificial intelligence is no exception. Along with Genomic Medicine and Biometrics, the UAE is pioneering AI use within and alongside other advanced technologies, stemming from the government's work in activating their strategy for Fourth Industrial Revolution (4IR). We are seeing exciting developments in augmented learning, personalised medicine, economic security (such as blockchain technology), smart cities, government e-services, 3D printing, and even space exploration. It's no secret that Dubai is earmarked to be a leading smart city, with investment into numerous technologies to set it on that course, and a stated aim by the UAE Government to make it a leading global hub and an open lab for the Fourth Industrial Revolution. The speed at which it is being implemented is almost unparalleled, and the opportunities that sit alongside that are vast.

It may seem from a dog's friendly demeanour and ability to perform tricks that they are smarter than most animals but a new study suggests this is not the case. Psychologists from Exeter and Canterbury University examined the cognitive abilities of'man's best friend' when compared with other animals - including cats. Experts concluded that canines do not possess particularly higher intelligence than their feline rivals, as well as a number of other creatures. Researchers used data on observations of the behaviour of dogs, cats, wolves and chimpanzees to see if pooches possessed any specific special skills. They found that dog's mental faculties are not exceptional compared to other species and were even bested in many categories.

Programming a robot to perform tasks requires training that is beyond the skill level of most individuals. To address this issue, we focus on developing a method that identifies keywords used to convey task knowledge among people and a framework that uses these keywords as conditions for knowledge acquisition by the robot learner. The methodology includes generalizing task modeling and providing a robot learner the ability to learn and improve its skills through accumulated experience gained from interaction with humans. More specifically, the aim of this research addresses the issues of knowledge encoding, acquisition, and retrieval through interactive instance-based learning (IIBL). In interaction studies, the benefit of using such a robot learner is in promoting social behaviors that results from the participant taking on an active role as teacher. Our recent experiment with 33 participants, including 19 typically developing children, and a pilot study with two children with autism spectrum disorder showed that IIBL provides a framework for designing an effective robot learner, and that the robot learner successfully increases the amount of social interactions initiated by the participants.