As the popularity of large language models (LLMs) soars across various applications, ensuring their alignment with human values has become a paramount concern. In particular, given that LLMs have great potential to serve as general-purpose AI assistants in daily life, their subtly unethical suggestions become a serious and real concern. Tackling the challenge of automatically testing and repairing unethical suggestions is thus demanding. This paper introduces the first framework for testing and repairing unethical suggestions made by LLMs. We first propose ETHICSSUITE, a test suite that presents complex, contextualized, and realistic moral scenarios to test LLMs. We then propose a novel suggest-critic-reflect (SCR) process, serving as an automated test oracle to detect unethical suggestions. We recast deciding if LLMs yield unethical suggestions (a hard problem; often requiring human expertise and costly to decide) into a PCR task that can be automatically checked for violation. Moreover, we propose a novel on-the-fly (OTF) repairing scheme that repairs unethical suggestions made by LLMs in real-time. The OTF scheme is applicable to LLMs in a black-box API setting with moderate cost. With ETHICSSUITE, our study on seven popular LLMs (e.g., ChatGPT, GPT-4) uncovers in total 109,824 unethical suggestions. We apply our OTF scheme on two LLMs (Llama-13B and ChatGPT), which generates valid repair to a considerable amount of unethical ones, paving the way for more ethically conscious LLMs.

Acceleration of large language model pre-training is a critical issue in present NLP research. In this paper, we focus on speeding up pre-training by progressively growing from a small Transformer structure to a large one. There are two main research problems related to progressive growth: growth schedule and growth operator. For growth schedule, existing work has explored multi-stage expansion of depth and feedforward layers. However, the impact of each dimension on the schedule's efficiency is still an open question. For growth operator, existing work relies on the initialization of new weights to inherit knowledge, and achieve only non-strict function preservation, limiting further optimization of training dynamics. To address these issues, we propose Masked Structural Growth (MSG), including growth schedules involving all possible dimensions and strictly function-preserving growth operators that is independent of the initialization of new weights. Experiments show that MSG is significantly faster than related work: we achieve a speed-up of 80% for Bert-base and 120% for Bert-large pre-training. Moreover, MSG is able to improve fine-tuning performances at the same time.

Recently, DeepNorm scales Transformers into extremely deep (i.e., 1000 layers) and reveals the promising potential of deep scaling. To stabilize the training of deep models, DeepNorm (Wang et al., 2022) attempts to constrain the model update to a constant value. Although applying such a constraint can benefit the early stage of model training, it may lead to undertrained models during the whole training procedure. In this paper, we propose BranchNorm, which dynamically rescales the non-residual branch of Transformer in accordance with the training period. BranchNorm not only theoretically stabilizes the training with smooth gradient norms at the early stage, but also encourages better convergence in the subsequent training stage. Experiment results on multiple translation tasks demonstrate that BranchNorm achieves a better trade-off between training stability and converge performance.

As large language models (LLMs) have become the norm in NLP, demonstrating good performance in generation and reasoning tasks, one of its most fatal disadvantages is the lack of factual correctness. Generating unfactual texts not only leads to lower performances but also degrades the trust and validity of their applications. Chain-of-Thought (CoT) prompting improves trust and model performance on complex reasoning tasks by generating interpretable reasoning chains, but still suffers from factuality concerns in knowledge-intensive tasks. In this paper, we propose the Verify-and-Edit framework for CoT prompting, which seeks to increase prediction factuality by post-editing reasoning chains according to external knowledge. Building on top of GPT-3, our framework lead to accuracy improvements in multiple open-domain question-answering tasks.

It due to their exceptional versatility in various is also the first released LLM of the Dandelion natural language proccessing tasks such as code Project. Our Panda LLM model has been writing and article editing, making them ubiquitous trained on Chinese-Wiki-2019, Chinese-News-in various industries and significantly enhancing 2016, Chinese-Baike-2018, Chinese-Webtext-2019 people's productivity (Ding et al., 2022; Zhao et al., and Translation-2019 (Xu, 2019) and COIG 2023). However, there are limitations to current datasets (Zhang et al., 2023) with instructiontuning off-the-shelf instruction-following large language (Wei et al., 2021) based on the LLaMA models, including the lack of trustworthiness in model (Touvron et al., 2023). Anticipated future releases generated results, lack of transparency in the model include progressively larger models such as used which raises concerns about data security, and Panda-13B and Panda-33B, with expected release the unknown training recipe, making it difficult to dates in the near future. Equal contribution, order decided by coin flip.

The artificial intelligence race is already producing losers. On Tuesday, education companies trading on the London and New York stock exchanges saw hundreds of millions wiped from their valuations after Chegg, a US firm that provides online help to students for writing and maths work, said ChatGPT was affecting customer growth. The firm said it had seen a "significant spike" in students using the technology, and withdrew its profits guidance for the rest of the year, warning revenues had already been hit. It shares almost halved in value. The ripples were felt in London, where education giant Pearson's stock closed down 15%.

As the buzz around ChatGPT and other generative AI increases, so has scammers' interest in the tech. In a new report published by Meta, the company says it's seen a sharp uptick in malware disguised as ChatGPT and similar AI software. In a statement, the company said that since March of 2023 alone, its researchers have discovered "ten malware families using ChatGPT and other similar themes to compromise accounts across the internet" and that it's blocked more than 1,000 malicious links from its platform. According to Meta, the scams often involve mobile apps or browser extensions posing as ChatGPT tools. And while in some cases the tools do offer some ChatGPT functionality, their real purpose is to steal their users' account credentials.

'The Five' co-hosts weigh in on the creator of ChatGPT raising'major concerns' regarding the implications of how artificial intelligence could change society. ChatGPT, the AI generative chatbot, has been trending as a helpful tool for many things day-to-day. As users play around with this artificial intelligence tool, some have used it to help plan out future travel. TikToker Madison Rolley, who posts regularly on the social media platform about budget travel, shared a video on April 12 explaining how she used ChatGPT to map out her next trip to Europe. The video went viral, with more than 250,000 viewers interested in how to use the bot for travel advice.

Tom Newhouse, vice president of Convergence Media, discusses the potential impact of artificial intelligence on elections after an RNC AI ad garnered attention. The Future of Life Institute has issued a petition to pause the development of GPT-5 and similar Large Language Models (LLMs). Their anxieties are understandable, but I believe they are much overblown. I've heard similar fears related to the advent of Artificial General Intelligence expressed off and on since I introduced the term AGI in 2005, but I think a pause would be a badly wrong move in the current situation for several reasons. Let me first emphasize something that's been mostly forgotten in the panic: Large Language Models can't become Artificial General Intelligences.

Fox News contributor Joe Concha joins'Fox & Friends First' to discuss Elon Musk's warning that artificial intelligence could threaten elections and his concerns on the declining birth rate. Artificial intelligence has become wildly popular for many Americans, but people over the age of 45 are trailing those younger than them on AI familiarity, a Fox News poll shows. Fifty-eight percent of registered voters over the age of 45 who were surveyed for the poll say they are not familiar with AI technology such as OpenAI's ChatGPT. Only 41% of registered voters over 45 reported they are familiar with the technology. The figures stand in stark contrast to younger Americans, with a whopping 65% of registered voters under the age of 45 reporting they are familiar with AI tech, such as ChatGPT.