Generative AI (GenAI) has revolutionized content generation, offering transformative capabilities for improving language coherence, readability, and overall quality. This manuscript explores the application of qualitative, quantitative, and mixed-methods research approaches to evaluate the performance of GenAI models in enhancing scientific writing. Using a hypothetical use case involving a collaborative medical imaging manuscript, we demonstrate how each method provides unique insights into the impact of GenAI. Qualitative methods gather in-depth feedback from expert reviewers, analyzing their responses using thematic analysis tools to capture nuanced improvements and identify limitations. Quantitative approaches employ automated metrics such as BLEU, ROUGE, and readability scores, as well as user surveys, to objectively measure improvements in coherence, fluency, and structure. Mixed-methods research integrates these strengths, combining statistical evaluations with detailed qualitative insights to provide a comprehensive assessment. These research methods enable quantifying improvement levels in GenAI-generated content, addressing critical aspects of linguistic quality and technical accuracy. They also offer a robust framework for benchmarking GenAI tools against traditional editing processes, ensuring the reliability and effectiveness of these technologies. By leveraging these methodologies, researchers can evaluate the performance boost driven by GenAI, refine its applications, and guide its responsible adoption in high-stakes domains like healthcare and scientific research. This work underscores the importance of rigorous evaluation frameworks for advancing trust and innovation in GenAI.

Peptides, short chains of amino acids, interact with target proteins, making them a unique class of protein-based therapeutics for treating human diseases. Recently, deep generative models have shown great promise in peptide generation. However, several challenges remain in designing effective peptide binders. First, not all residues contribute equally to peptide-target interactions. Second, the generated peptides must adopt valid geometries due to the constraints of peptide bonds. Third, realistic tasks for peptide drug development are still lacking. To address these challenges, we introduce PepHAR, a hot-spot-driven autoregressive generative model for designing peptides targeting specific proteins. Building on the observation that certain hot spot residues have higher interaction potentials, we first use an energy-based density model to fit and sample these key residues. Next, to ensure proper peptide geometry, we autoregressively extend peptide fragments by estimating dihedral angles between residue frames. Finally, we apply an optimization process to iteratively refine fragment assembly, ensuring correct peptide structures. By combining hot spot sampling with fragment-based extension, our approach enables de novo peptide design tailored to a target protein and allows the incorporation of key hot spot residues into peptide scaffolds. Extensive experiments, including peptide design and peptide scaffold generation, demonstrate the strong potential of PepHAR in computational peptide binder design.

This study explores the application of generative AI (GenAI) within manual exploitation and privilege escalation tasks in Linux-based penetration testing environments, two areas critical to comprehensive cybersecurity assessments. Building on previous research into the role of GenAI in the ethical hacking lifecycle, this paper presents a hands-on experimental analysis conducted in a controlled virtual setup to evaluate the utility of GenAI in supporting these crucial, often manual, tasks. Our findings demonstrate that GenAI can streamline processes, such as identifying potential attack vectors and parsing complex outputs for sensitive data during privilege escalation. The study also identifies key benefits and challenges associated with GenAI, including enhanced efficiency and scalability, alongside ethical concerns related to data privacy, unintended discovery of vulnerabilities, and potential for misuse. This work contributes to the growing field of AI-assisted cybersecurity by emphasising the importance of human-AI collaboration, especially in contexts requiring careful decision-making, rather than the complete replacement of human input.

NVIDIA has debuted a new experimental generative AI model, which it describes as "a Swiss Army knife for sound." The model called Foundational Generative Audio Transformer Opus 1, or Fugatto, can take commands from text prompts and use them to create audio or to modify existing music, voice and sound files. It was designed by a team of AI researchers from around the world, and NVIDIA says that made the model's "multi-accent and multilingual capabilities stronger." "We wanted to create a model that understands and generates sound like humans do," said Rafael Valle, one of the researchers behind the project and a manager of applied audio research at NVIDIA. The company listed some possible real-world scenarios wherein Fugatto could be of use in its announcement. Music producers, it suggested, could use the technology to quickly generate a prototype for a song idea, which they can then easily edit to try out different styles, voices and instruments.

Meta will make its generative artificial intelligence (AI) models available to the United States' government, the tech giant has announced, in a controversial move that raises a moral dilemma for everyone who uses the software. Meta last week revealed it would make the models, known as Llama, available to government agencies, "including those that are working on defence and national security applications, and private sector partners supporting their work". The decision appears to contravene Meta's own policy which lists a range of prohibited uses for Llama, including "[m]ilitary, warfare, nuclear industries or applications" as well as espionage, terrorism, human trafficking and exploitation or harm to children. Meta's exception also reportedly applies to similar national security agencies in the United Kingdom, Canada, Australia and New Zealand. It came just three days after Reuters revealed China has reworked Llama for its own military purposes.

Recent advances in Diffusion Models have enabled the generation of images from text, with powerful closed-source models like DALL-E and Midjourney leading the way. However, open-source alternatives, such as StabilityAI's Stable Diffusion, offer comparable capabilities. These open-source models, hosted on Hugging Face, come equipped with ethical filter protections designed to prevent the generation of explicit images. This paper reveals first their limitations and then presents a novel text-based safety filter that outperforms existing solutions. Our research is driven by the critical need to address the misuse of AI-generated content, especially in the context of information warfare. DiffGuard enhances filtering efficacy, achieving a performance that surpasses the best existing filters by over 14%.

This paper presents a critical examination of current approaches to replicating OpenAI's O1 model capabilities, with particular focus on the widespread but often undisclosed use of knowledge distillation techniques. While our previous work (Part 1 (Qin et al., 2024)) explored the fundamental technical path to O1 replication, this study reveals how simple distillation from O1's API, combined with supervised fine-tuning, can achieve superior performance on complex mathematical reasoning tasks. Through extensive experiments, we show that a base model fine-tuned on simply tens of thousands of samples O1-distilled long-thought chains outperforms O1-preview on the American Invitational Mathematics Examination (AIME) with minimal technical complexity. Moreover, our investigation extends beyond mathematical reasoning to explore the generalization capabilities of O1-distilled models across diverse tasks: hallucination, safety and open-domain QA. Notably, despite training only on mathematical problem-solving data, our models demonstrated strong generalization to open-ended QA tasks and became significantly less susceptible to sycophancy after fine-tuning. We deliberately make this finding public to promote transparency in AI research and to challenge the current trend of obscured technical claims in the field. Our work includes: (1) A detailed technical exposition of the distillation process and its effectiveness, (2) A comprehensive benchmark framework for evaluating and categorizing O1 replication attempts based on their technical transparency and reproducibility, (3) A critical discussion of the limitations and potential risks of over-relying on distillation approaches, our analysis culminates in a crucial "bitter lesson": while the pursuit of more capable AI systems is important, the development of researchers grounded in firstprinciples thinking is paramount. This educational imperative represents not just a technical consideration, but a fundamental human mission that will shape the future of AI innovation.

With the growing popularity of generative AI for images, video, and music, we witnessed models rapidly improve in quality and performance. However, not much attention is paid towards enabling AI's ability to "be creative". In this study, we implemented and simulated the systems model of creativity (proposed by Csikszentmihalyi) using virtual agents utilizing large language models (LLMs) and text prompts. For comparison, the simulations were conducted with the "virtual artists" being: 1)isolated and 2)placed in a multi-agent system. Both scenarios were compared by analyzing the variations and overall "creativity" in the generated artifacts (measured via a user study and LLM). Our results suggest that the generative agents may perform better in the framework of the systems model of creativity.

Despite remarkable progress in multi-modal AI research, there is a salient domain in which modern AI continues to lag considerably behind even human children: the reliable deployment of logical operators. Here, we examine three forms of logical operators: relations, negations, and discrete numbers. We asked human respondents (N=178 in total) to evaluate images generated by a state-of-the-art image-generating AI (DALL-E 3) prompted with these `logical probes', and find that none reliably produce human agreement scores greater than 50\%. The negation probes and numbers (beyond 3) fail most frequently. In a 4th experiment, we assess a `grounded diffusion' pipeline that leverages targeted prompt engineering and structured intermediate representations for greater compositional control, but find its performance is judged even worse than that of DALL-E 3 across prompts. To provide further clarity on potential sources of success and failure in these text-to-image systems, we supplement our 4 core experiments with multiple auxiliary analyses and schematic diagrams, directly quantifying, for example, the relationship between the N-gram frequency of relational prompts and the average match to generated images; the success rates for 3 different prompt modification strategies in the rendering of negation prompts; and the scalar variability / ratio dependence (`approximate numeracy') of prompts involving integers. We conclude by discussing the limitations inherent to `grounded' multimodal learning systems whose grounding relies heavily on vector-based semantics (e.g. DALL-E 3), or under-specified syntactical constraints (e.g. `grounded diffusion'), and propose minimal modifications (inspired by development, based in imagery) that could help to bridge the lingering compositional gap between scale and structure. All data and code is available at https://github.com/ColinConwell/T2I-Probology

The manual assessment and grading of student writing is a time-consuming yet critical task for teachers. Recent developments in generative AI, such as large language models, offer potential solutions to facilitate essay-scoring tasks for teachers. In our study, we evaluate the performance and reliability of both open-source and closed-source LLMs in assessing German student essays, comparing their evaluations to those of 37 teachers across 10 pre-defined criteria (i.e., plot logic, expression). A corpus of 20 real-world essays from Year 7 and 8 students was analyzed using five LLMs: GPT-3.5, GPT-4, o1, LLaMA 3-70B, and Mixtral 8x7B, aiming to provide in-depth insights into LLMs' scoring capabilities. Closed-source GPT models outperform open-source models in both internal consistency and alignment with human ratings, particularly excelling in language-related criteria. The novel o1 model outperforms all other LLMs, achieving Spearman's $r = .74$ with human assessments in the overall score, and an internal consistency of $ICC=.80$. These findings indicate that LLM-based assessment can be a useful tool to reduce teacher workload by supporting the evaluation of essays, especially with regard to language-related criteria. However, due to their tendency for higher scores, the models require further refinement to better capture aspects of content quality.