One of the most stunning twists in the recent five-day crisis at ChatGPT creator OpenAI came when some 95 percent of the company's hundreds of employees threatened to quit. The staff planned to follow CEO Sam Altman to develop successors to ChatGPT at Microsoft instead. The threat appeared to mark a turning point in Altman's ultimately successful attempt to return to OpenAI--it was also a scenario that businesses have the legal power to block in most US states. California, home to OpenAI's San Francisco HQ, is one of a handful states that bar the enforcement of noncompete agreements in employment contracts, which can forbid employees from hopping jobs to a competitor, often for years. That picture is now set to change, as a raft of new legislation aims to make more places like California.

Rosário said the chatbot processed a 250-character command and took some 15 seconds to employ its algorithmic magic and spit out a policy -- a process that would normally take him about three days. The result, he said, showcased how artificial intelligence can be a useful tool for optimizing and improving public service. Yet Brazil's first ChatGPT-crafted law has launched the South American nation into a debate ringing across the globe: As artificial intelligence takes the world by storm, is society gearing toward a future where automation replaces humans?

Worked examples (solutions to typical programming problems presented as a source code in a certain language and are used to explain the topics from a programming class) are among the most popular types of learning content in programming classes. Most approaches and tools for presenting these examples to students are based on line-by-line explanations of the example code. However, instructors rarely have time to provide line-by-line explanations for a large number of examples typically used in a programming class. In this paper, we explore and assess a human-AI collaboration approach to authoring worked examples for Java programming. We introduce an authoring system for creating Java worked examples that generates a starting version of code explanations and presents it to the instructor to edit if necessary. We also present a study that assesses the quality of explanations created with this approach.

Large language models (LLMs) have demonstrated impressive capabilities in storing and recalling factual knowledge, but also in adapting to novel in-context information. Yet, the mechanisms underlying their in-context grounding remain unknown, especially in situations where in-context information contradicts factual knowledge embedded in the parameters. This is critical for retrieval-augmented generation methods, which enrich the context with up-to-date information, hoping that grounding can rectify the outdated parametric knowledge. In this study, we introduce Fakepedia, a counterfactual dataset designed to evaluate grounding abilities when the parametric knowledge clashes with the in-context information. We benchmark various LLMs with Fakepedia and discover that GPT-4-turbo has a strong preference for its parametric knowledge. Mistral-7B, on the contrary, is the model that most robustly chooses the grounded answer. Then, we conduct causal mediation analysis on LLM components when answering Fakepedia queries. We demonstrate that inspection of the computational graph alone can predict LLM grounding with 92.8% accuracy, especially because few MLPs in the Transformer can predict non-grounded behavior. Our results, together with existing findings about factual recall mechanisms, provide a coherent narrative of how grounding and factual recall mechanisms interact within LLMs.

As Artificial Intelligence (AI) technologies continue to gain traction in the modern-day world, they ultimately pose an immediate threat to current cybersecurity systems via exploitative methods. Prompt engineering is a relatively new field that explores various prompt designs that can hijack large language models (LLMs). If used by an unethical attacker, it can enable an AI system to offer malicious insights and code to them. In this paper, an enhanced intrusion detection system (IDS) that utilizes machine learning (ML) and hyperparameter tuning is explored, which can improve a model's performance in terms of accuracy and efficacy. Ultimately, this improved system can be used to combat the attacks made by unethical hackers. A standard IDS is solely configured with pre-configured rules and patterns; however, with the utilization of machine learning, implicit and different patterns can be generated through the models' hyperparameter settings and parameters. In addition, the IDS will be equipped with multiple datasets so that the accuracy of the models improves. We evaluate the performance of multiple ML models and their respective hyperparameter settings through various metrics to compare their results to other models and past research work. The results of the proposed multi-dataset integration method yielded an accuracy score of 99.9% when equipped with the XGBoost and random forest classifiers and RandomizedSearchCV hyperparameter technique.

The advancement of Large Language Models (LLMs) has brought substantial attention to the Chain of Thought (CoT) approach Wei et al. [2022a], primarily due to its ability to enhance the capability of LLMs on tasks requiring complex reasoning. Moreover, the significance of CoT approaches extends to the application of LLMs for multi-modal tasks, such as multi-modal question answering. However, the selection of optimal CoT demonstration examples in multi-modal reasoning for LLMs remains less explored for LLMs due to the inherent complexity of multi-modal examples. In this paper, we introduce a novel approach that addresses this challenge by using retrieval mechanisms to dynamically and automatically select demonstration examples based on cross-modal similarities. This method aims to refine the CoT reasoning process in multi-modal scenarios via informing LLMs with more relevant and informative examples. Furthermore, we employ a stratified sampling method categorising demonstration examples into groups based on their types and retrieving examples from different groups respectively to promote the diversity of demonstration examples. Through a series of experiments, we demonstrate that our approach significantly improves the performance of LLMs, achieving state-of-the-art results in multi-modal reasoning tasks. Specifically, our methods demonstrate significant advancements on the ScienceQA dataset. While our method based on ChatGPT outperforms the Chameleon (ChatGPT) by 2.74% with an accuracy of 82.67%, the GPT4-based approach surpasses the Chameleon (GPT-4) by 0.89%, achieving 87.43% on accuracy under the same setting.

We propose Neural Priming, a technique for adapting large pretrained models to distribution shifts and downstream tasks given few or no labeled examples. Presented with class names or unlabeled test samples, Neural Priming enables the model to recall and conditions its parameters on relevant data seen throughout pretraining, thereby priming it for the test distribution. Neural Priming can be performed at test time, even for pretraining datasets as large as LAION-2B. Performing lightweight updates on the recalled data significantly improves accuracy across a variety of distribution shift and transfer learning benchmarks. Concretely, in the zero-shot setting, we see a 2.45% improvement in accuracy on ImageNet and 3.81% accuracy improvement on average across standard transfer learning benchmarks. Further, using Neural Priming at inference to adapt to distribution shift, we see a 1.41% accuracy improvement on ImageNetV2. These results demonstrate the effectiveness of Neural Priming in addressing the challenge of limited labeled data and changing distributions. Code is available at github.com/RAIVNLab/neural-priming.

With the increasing use of large-language models (LLMs) like ChatGPT, watermarking has emerged as a promising approach for tracing machine-generated content. However, research on LLM watermarking often relies on simple perplexity or diversity-based measures to assess the quality of watermarked text, which can mask important limitations in watermarking. Here we introduce two new easy-to-use methods for evaluating watermarking algorithms for LLMs: 1) evaluation by LLM-judger with specific guidelines; and 2) binary classification on text embeddings to distinguish between watermarked and unwatermarked text. We apply these methods to characterize the effectiveness of current watermarking techniques. Our experiments, conducted across various datasets, reveal that current watermarking methods are detectable by even simple classifiers, challenging the notion of watermarking subtlety. We also found, through the LLM judger, that watermarking impacts text quality, especially in degrading the coherence and depth of the response. Our findings underscore the trade-off between watermark robustness and text quality and highlight the importance of having more informative metrics to assess watermarking quality.

While intelligent virtual assistants like Siri, Alexa, and Google Assistant have become ubiquitous in modern life, they still face limitations in their ability to follow multi-step instructions and accomplish complex goals articulated in natural language. However, recent breakthroughs in large language models (LLMs) show promise for overcoming existing barriers by enhancing natural language processing and reasoning capabilities. Though promising, applying LLMs to create more advanced virtual assistants still faces challenges like ensuring robust performance and handling variability in real-world user commands. This paper proposes a novel LLM-based virtual assistant that can automatically perform multi-step operations within mobile apps based on high-level user requests. The system represents an advance in assistants by providing an end-to-end solution for parsing instructions, reasoning about goals, and executing actions. LLM-based Process Automation (LLMPA) has modules for decomposing instructions, generating descriptions, detecting interface elements, predicting next actions, and error checking. Experiments demonstrate the system completing complex mobile operation tasks in Alipay based on natural language instructions. This showcases how large language models can enable automated assistants to accomplish real-world tasks. The main contributions are the novel LLMPA architecture optimized for app process automation, the methodology for applying LLMs to mobile apps, and demonstrations of multi-step task completion in a real-world environment. Notably, this work represents the first real-world deployment and extensive evaluation of a large language model-based virtual assistant in a widely used mobile application with an enormous user base numbering in the hundreds of millions.

When a damaging earthquake occurs, immediate information about casualties is critical for time-sensitive decision-making by emergency response and aid agencies in the first hours and days. Systems such as Prompt Assessment of Global Earthquakes for Response (PAGER) by the U.S. Geological Survey (USGS) were developed to provide a forecast within about 30 minutes of any significant earthquake globally. Traditional systems for estimating human loss in disasters often depend on manually collected early casualty reports from global media, a process that's labor-intensive and slow with notable time delays. Recently, some systems have employed keyword matching and topic modeling to extract relevant information from social media. However, these methods struggle with the complex semantics in multilingual texts and the challenge of interpreting ever-changing, often conflicting reports of death and injury numbers from various unverified sources on social media platforms. In this work, we introduce an end-to-end framework to significantly improve the timeliness and accuracy of global earthquake-induced human loss forecasting using multi-lingual, crowdsourced social media. Our framework integrates (1) a hierarchical casualty extraction model built upon large language models, prompt design, and few-shot learning to retrieve quantitative human loss claims from social media, (2) a physical constraint-aware, dynamic-truth discovery model that discovers the truthful human loss from massive noisy and potentially conflicting human loss claims, and (3) a Bayesian updating loss projection model that dynamically updates the final loss estimation using discovered truths. We test the framework in real-time on a series of global earthquake events in 2021 and 2022 and show that our framework streamlines casualty data retrieval, achieving speed and accuracy comparable to manual methods by USGS.