This is not the first time that I’ve written about AI creativity, and I doubt that it will be the last. It’s a question that comes up repeatedly, and that is very much in the current mind, with events like the strikes by the… Read More »Creativity Isn’t Just Remixing Mike Loukides AI & ML – Radar
The structure of Ghostbuster, our new state-of-the-art method for detecting AI-generated text.
Large language models like ChatGPT write impressively well—so well, in fact, that they’ve become a problem. Students have begun using these models to ghostwrite assignments, leading some schools to ban ChatGPT. In addition, these models are also prone to producing text with factual errors, so wary readers may want to know if generative AI tools have been used to ghostwrite news articles or other sources before trusting them.
What can teachers and consumers do? Existing tools to detect AI-generated text sometimes do poorly on data that differs from what they were trained on. In addition, if these models falsely classify real human writing as AI-generated, they can jeopardize students whose genuine work is called into question.
Our recent paper introduces Ghostbuster, a state-of-the-art method for detecting AI-generated text. Ghostbuster works by finding the probability of generating each token in a document under several weaker language models, then combining functions based on these probabilities as input to a final classifier. Ghostbuster doesn’t need to know what model was used to generate a document, nor the probability of generating the document under that specific model. This property makes Ghostbuster particularly useful for detecting text potentially generated by an unknown model or a black-box model, such as the popular commercial models ChatGPT and Claude, for which probabilities aren’t available. We’re particularly interested in ensuring that Ghostbuster generalizes well, so we evaluated across a range of ways that text could be generated, including different domains (using newly collected datasets of essays, news, and stories), language models, or prompts.
In an era dominated by AI advancements, distinguishing between human and machine-generated content, especially in scientific publications, has become increasingly pressing. This paper addresses this concern head-on, proposing a robust solution to identify and differentiate between human and AI-generated writing accurately for chemistry papers.… Read More »A New Research Paper Introduces a Machine-Learning Tool that can Easily Spot when Chemistry Papers are Written Using the Chatbot ChatGPT Madhur Garg Artificial Intelligence Category – MarkTechPost
Medical data extraction, analysis, and interpretation from unstructured clinical literature are included in the emerging discipline of clinical natural language processing (NLP). Even with its importance, particular difficulties arise while developing methodologies for clinical NLP. For instance, clinical texts might confuse ordinary NLP models… Read More »Can Synthetic Clinical Text Generation Revolutionize Clinical NLP Tasks? Meet ClinGen: An AI Model that Involves Clinical Knowledge Extraction and Context-Informed LLM Prompting Aneesh Tickoo Artificial Intelligence Category – MarkTechPost
Asymmetric Certified Robustness via Feature-Convex Neural Networks
TLDR: We propose the asymmetric certified robustness problem, which requires certified robustness for only one class and reflects real-world adversarial scenarios. This focused setting allows us to introduce feature-convex classifiers, which produce closed-form and deterministic certified radii on the order of milliseconds.
Figure 1. Illustration of feature-convex classifiers and their certification for sensitive-class inputs. This architecture composes a Lipschitz-continuous feature map $varphi$ with a learned convex function $g$. Since $g$ is convex, it is globally underapproximated by its tangent plane at $varphi(x)$, yielding certified norm balls in the feature space. Lipschitzness of $varphi$ then yields appropriately scaled certificates in the original input space.
Despite their widespread usage, deep learning classifiers are acutely vulnerable to adversarial examples: small, human-imperceptible image perturbations that fool machine learning models into misclassifying the modified input. This weakness severely undermines the reliability of safety-critical processes that incorporate machine learning. Many empirical defenses against adversarial perturbations have been proposed—often only to be later defeated by stronger attack strategies. We therefore focus on certifiably robust classifiers, which provide a mathematical guarantee that their prediction will remain constant for an $ell_p$-norm ball around an input.
Conventional certified robustness methods incur a range of drawbacks, including nondeterminism, slow execution, poor scaling, and certification against only one attack norm. We argue that these issues can be addressed by refining the certified robustness problem to be more aligned with practical adversarial settings.
Researchers from ETH Zurich explore simplifications in the design of deep Transformers, aiming to make them more robust and efficient. Modifications are proposed by combining signal propagation theory and empirical observations, enabling the removal of various components from standard transformer blocks without compromising training… Read More »Can Transformer Blocks Be Simplified Without Compromising Efficiency? This AI Paper from ETH Zurich Explores the Balance Between Design Complexity and Performance Adnan Hassan Artificial Intelligence Category – MarkTechPost
In software development, developers often face challenges when working with complex code or managing project issues efficiently. Finding the correct information and assistance in the workflow can be a hurdle. To address this, GitLab has introduced a new tool called Duo Chat, which aims… Read More »GitLab Introduces Duo Chat: A Conversational AI Tool for Productivity Niharika Singh Artificial Intelligence Category – MarkTechPost
In the realm of immersive experiences in mixed-reality scenarios, generating accurate and plausible full-body avatar motion has been a persistent challenge. Existing solutions relying on Head-Mounted Devices (HMDs) typically utilize limited input signals, such as head and hands 6-DoF (degrees of freedom). While recent… Read More »A New Microsoft AI Research Proposes HMD-NeMo: A New Approach that Addresses Plausible and Accurate Full Body Motion Generation Even When the Hands may be Only Partially Visible Pragati Jhunjhunwala Artificial Intelligence Category – MarkTechPost
Given the success of diffusion models in text-to-image generation, a surge of video generation techniques has emerged, showcasing interesting applications in this realm. Nevertheless, most video generation techniques often produce videos at the “shot-level,” enclosing only a few seconds and portraying a single scene.… Read More »Meet SEINE: a Short-to-Long Video Diffusion Model for High-Quality Extended Videos with Smooth and Creative Transitions Between Scenes Daniele Lorenzi Artificial Intelligence Category – MarkTechPost
NVIDIA today unveiled at SC23 the next wave of technologies that will lift scientific and industrial research centers worldwide to new levels of performance and energy efficiency. “NVIDIA hardware and software innovations are creating a new class of AI supercomputers,” said Ian Buck, vice… Read More »New Class of Accelerated, Efficient AI Systems Mark the Next Era of Supercomputing Rick Merritt – Archives Page 1 | NVIDIA Blog