Once again, team Engadget has set up shop in Las Vegas for CES, living out of suitcases so that we can scour the massive show floor and occasionally injure ourselves in the process. For CES 2024, we expected to see AI everywhere, and we were not disappointed. We saw more than a few laptops with AI-powered chips inside, not to mention multiple references to Microsoft's Copilot assistant. Volkswagen built ChatGPT into its in-car system, while BMW teamed up with Amazon to improve its own in-car assistant. Qualcomm announced an AI Snapdragon chip.

You probably already own an AI PC. In the past few months, Intel and PC makers have beat the drum of the AI PC loudly and in concert with AMD, Intel, and Qualcomm. It's no secret that "AI" is the new "metaverse" -- you know, that thing that everyone was talking up a few years ago -- and executives and investors alike want to use AI to boost sales and stock prices. And it's true that AI does depend on the NPUs found in chips like Intel's Core Ultra, the brand that Intel is positioning as synonymous with on-chip AI. The same goes for AMD's Ryzen 8000 series -- which beat Intel to the desktop with an NPU -- as well as Qualcomm's Snapdragon X Elite.

And because a LLM doesn't "know" when it's quoting from training data, there's no obvious way to prevent the behavior. I spoke with Florian Tramèr, a prominent AI-security researcher and co-author of some of the above studies. It's "an extremely tricky problem to study," he told me. "It's very, very hard to pin down a good definition of memorization." One way to understand the concept is to think of an LLM as an enormous decision tree in which each node is an English word. From a given starting word, an LLM chooses the next word from the entire English vocabulary.

This is a staggeringly difficult problem to solve as current AI cannot look at a blank canvas and create something original,

When Walmart announced it would be holding a CES keynote for the first time, we were admittedly a little skeptical. Now it all makes sense, though: America's largest retailer came to CES 2024 in Las Vegas to talk about AI. In a joint announcement on Tuesday, the company said that it's teaming up with Microsoft to build what it bills as AI-powered shopping experiences. In his keynote, Walmart CEO Doug McMillon described how the integration of AI across its website and apps will be used to study shopper behavior and suggest future purchases. As you might expect, given Microsoft's involvement, the artificial intelligence underpinning these experiences will be powered by large language models made available through this partnership with Microsoft.

Question generation (QG) is a natural language processing task with an abundance of potential benefits and use cases in the educational domain. In order for this potential to be realized, QG systems must be designed and validated with pedagogical needs in mind. However, little research has assessed or designed QG approaches with the input from real teachers or students. This paper applies a large language model-based QG approach where questions are generated with learning goals derived from Bloom's taxonomy. The automatically generated questions are used in multiple experiments designed to assess how teachers use them in practice. The results demonstrate that teachers prefer to write quizzes with automatically generated questions, and that such quizzes have no loss in quality compared to handwritten versions. Further, several metrics indicate that automatically generated questions can even improve the quality of the quizzes created, showing the promise for large scale use of QG in the classroom setting.

Large Language Models (LLMs) have been widely employed in programming language analysis to enhance human productivity. Yet, their reliability can be compromised by various code distribution shifts, leading to inconsistent outputs. While probabilistic methods are known to mitigate such impact through uncertainty calibration and estimation, their efficacy in the language domain remains underexplored compared to their application in image-based tasks. In this work, we first introduce a large-scale benchmark dataset, incorporating three realistic patterns of code distribution shifts at varying intensities. Then we thoroughly investigate state-of-the-art probabilistic methods applied to CodeLlama using these shifted code snippets. We observe that these methods generally improve the uncertainty awareness of CodeLlama, with increased calibration quality and higher uncertainty estimation~(UE) precision. However, our study further reveals varied performance dynamics across different criteria (e.g., calibration error vs misclassification detection) and trade-off between efficacy and efficiency, highlighting necessary methodological selection tailored to specific contexts.

Software engineers are increasingly adding semantic search capabilities to applications using a strategy known as Retrieval Augmented Generation (RAG). A RAG system involves finding documents that semantically match a query and then passing the documents to a large language model (LLM) such as ChatGPT to extract the right answer using an LLM. RAG systems aim to: a) reduce the problem of hallucinated responses from LLMs, b) link sources/references to generated responses, and c) remove the need for annotating documents with meta-data. However, RAG systems suffer from limitations inherent to information retrieval systems and from reliance on LLMs. In this paper, we present an experience report on the failure points of RAG systems from three case studies from separate domains: research, education, and biomedical. We share the lessons learned and present 7 failure points to consider when designing a RAG system. The two key takeaways arising from our work are: 1) validation of a RAG system is only feasible during operation, and 2) the robustness of a RAG system evolves rather than designed in at the start. We conclude with a list of potential research directions on RAG systems for the software engineering community.

Self-supervised learning has emerged as a highly effective approach in the fields of natural language processing and computer vision. It is also applicable to brain signals such as electroencephalography (EEG) data, given the abundance of available unlabeled data that exist in a wide spectrum of real-world medical applications ranging from seizure detection to wave analysis. The existing works leveraging self-supervised learning on EEG modeling mainly focus on pretraining upon each individual dataset corresponding to a single downstream task, which cannot leverage the power of abundant data, and they may derive sub-optimal solutions with a lack of generalization. Moreover, these methods rely on end-to-end model learning which is not easy for humans to understand. In this paper, we present a novel EEG foundation model, namely EEGFormer, pretrained on large-scale compound EEG data. The pretrained model cannot only learn universal representations on EEG signals with adaptable performance on various downstream tasks but also provide interpretable outcomes of the useful patterns within the data. To validate the effectiveness of our model, we extensively evaluate it on various downstream tasks and assess the performance under different transfer settings. Furthermore, we demonstrate how the learned model exhibits transferable anomaly detection performance and provides valuable interpretability of the acquired patterns via self-supervised learning.

Getting large language models (LLMs) to refuse to answer hostile toxicity questions is a core issue under the theme of LLMs security. Previous approaches have used prompts engineering to jailbreak LLMs and answer some toxicity questions. These approaches can easily fail after the model manufacturer makes additional fine-tuning to the model. To promote the further understanding of model jailbreaking by researchers, we are inspired by Representation Engineering to propose a jailbreaking method that does not require elaborate construction prompts, is not affected by model fine-tuning, and can be widely applied to any open-source LLMs in a pluggable manner. We have evaluated this method on multiple mainstream LLMs on carefully supplemented toxicity datasets, and the experimental results demonstrate the significant effectiveness of our approach. After being surprised by some interesting jailbreaking cases, we did extensive in-depth research to explore the techniques behind this method.