Hate speech is a harmful form of online expression, often manifesting as derogatory posts. It is a significant risk in digital environments. With the rise of Large Language Models (LLMs), there is concern about their potential to replicate hate speech patterns, given their training on vast amounts of unmoderated internet data. Understanding how LLMs respond to hate speech is crucial for their responsible deployment. However, the behaviour of LLMs towards hate speech has been limited compared. This paper investigates the reactions of seven state-of-the-art LLMs (LLaMA 2, Vicuna, LLaMA 3, Mistral, GPT-3.5, GPT-4, and Gemini Pro) to hate speech. Through qualitative analysis, we aim to reveal the spectrum of responses these models produce, highlighting their capacity to handle hate speech inputs. We also discuss strategies to mitigate hate speech generation by LLMs, particularly through fine-tuning and guideline guardrailing. Finally, we explore the models' responses to hate speech framed in politically correct language.

Text summarization tasks commonly employ Pre-trained Language Models (PLMs) to fit diverse standard datasets. While these PLMs excel in automatic evaluations, they frequently underperform in human evaluations, indicating a deviation between their generated summaries and human summarization preferences. This discrepancy is likely due to the low quality of fine-tuning datasets and the limited availability of high-quality human-annotated data that reflect true human preference. To address this challenge, we introduce a novel human summarization preference alignment framework AlignSum. This framework consists of three parts: Firstly, we construct a Data Pymarid with extractive, abstractive, and human-annotated summary data. Secondly, we conduct the Gaussian Resampling to remove summaries with extreme lengths. Finally, we implement the two-stage hierarchical fine-tuning with Data Pymarid after Gaussian Resampling. We apply AlignSum to PLMs on the human-annotated CNN/DailyMail and BBC XSum datasets. Experiments show that with AlignSum, PLMs like BART-Large surpass 175B GPT-3 in both automatic and human evaluations. This demonstrates that AlignSum significantly enhances the alignment of language models with human summarization preferences.

As large language models (LLMs) are applied across diverse domains, the ability to selectively unlearn specific information has become increasingly essential. For instance, LLMs are expected to provide confidential information to authorized internal users, such as employees or trusted partners, while withholding it from external users, including the general public and unauthorized entities. In response to this challenge, we propose a novel method termed ``in-context knowledge unlearning'', which enables the model to selectively forget information in test-time based on the context of the query. Our method fine-tunes pre-trained LLMs to enable prompt unlearning of target knowledge within the context, while preserving other knowledge. Experiments on the TOFU and AGE datasets using Llama2-7B/13B and Mistral-7B models show our method achieves up to 95% forgetting accuracy while retaining 80% of unrelated knowledge, significantly outperforming baselines in both in-domain and out-of-domain scenarios. Further investigation into the model's internal behavior revealed that while fine-tuned LLMs generate correct predictions in the middle layers and maintain them up to the final layer, they make the decision to forget at the last layer, i.e., ``LLMs pretend to forget''. Our findings offer valuable insights into enhancing the robustness of unlearning mechanisms in LLMs, setting a foundation for future research in the field.

ChatGPT and other state-of-the-art large language models (LLMs) are rapidly transforming multiple fields, offering powerful tools for a wide range of applications. These models, commonly trained on vast datasets, exhibit human-like text generation capabilities, making them useful for research tasks such as ideation, literature review, coding, drafting, and outreach. We conducted a study involving 13 astronomers at different career stages and research fields to explore LLM applications across diverse tasks over several months and to evaluate their performance in research-related activities. This work was accompanied by an anonymous survey assessing participants' experiences and attitudes towards LLMs. We provide a detailed analysis of the tasks attempted and the survey answers, along with specific output examples. Our findings highlight both the potential and limitations of LLMs in supporting research while also addressing general and research-specific ethical considerations. We conclude with a series of recommendations, emphasizing the need for researchers to complement LLMs with critical thinking and domain expertise, ensuring these tools serve as aids rather than substitutes for rigorous scientific inquiry.

Many applications involve reasoning about time durations before a critical event happens--also called time-to-event outcomes. When will a customer cancel a subscription, a coma patient wake up, or a convicted criminal reoffend? Time-to-event outcomes have been studied extensively within the field of survival analysis primarily by the statistical, medical, and reliability engineering communities, with textbooks already available in the 1970s and '80s. This monograph aims to provide a reasonably self-contained modern introduction to survival analysis. We focus on predicting time-to-event outcomes at the individual data point level with the help of neural networks. Our goal is to provide the reader with a working understanding of precisely what the basic time-to-event prediction problem is, how it differs from standard regression and classification, and how key "design patterns" have been used time after time to derive new time-to-event prediction models, from classical methods like the Cox proportional hazards model to modern deep learning approaches such as deep kernel Kaplan-Meier estimators and neural ordinary differential equation models. We further delve into two extensions of the basic time-to-event prediction setup: predicting which of several critical events will happen first along with the time until this earliest event happens (the competing risks setting), and predicting time-to-event outcomes given a time series that grows in length over time (the dynamic setting). We conclude with a discussion of a variety of topics such as fairness, causal reasoning, interpretability, and statistical guarantees. Our monograph comes with an accompanying code repository that implements every model and evaluation metric that we cover in detail.

The term "AI" is inescapable right now, spreading its way across every facet of every market, whether it actually makes sense or not. And as it turns out, the US Federal Trade Commission (FTC) is as sick of all this AI-related branding as everyone else. A recent regulatory push dubbed "Operation AI Comply" now has the FTC cracking down on some of the more notable AI implementations, including at least three alleged scams. A press release from last week detailed five new cases that the FTC is taking on, specifically targeting firms that sprinkle AI-related claims into their businesses. It quotes chairperson Lina M. Khan, who's been particularly proactive since her appointment in 2021: "Using AI tools to trick, mislead, or defraud people is illegal. The FTC's enforcement actions make clear that there is no AI exemption from the laws on the books. By cracking down on unfair or deceptive practices in these markets, FTC is ensuring that honest businesses and innovators can get a fair shot and consumers are being protected."

California Governor Gavin Newsom has vetoed what would have become one of the most comprehensive policies governing the safety of artificial intelligence in the U.S. The bill would've been among the first to hold AI developers accountable for any severe harm caused by their technologies. It drew fierce criticism from some prominent Democrats and major tech firms, including ChatGPT creator OpenAI and venture capital firm Andreessen Horowitz, who warned it could stall innovation in the state. Newsom described the legislation as "well-intentioned" but said in a statement that it would've applied "stringent standards to even the most basic functions." Regulation should be based on "empirical evidence and science," he said, pointing to his own executive order on AI and other bills he's signed that regulate the technology around known risks such as deepfakes. The debate around California's SB 1047 bill highlights the challenge that lawmakers around the world are facing in controlling the risks of AI while also supporting the emerging technology.

Backstage at one of the largest artificial intelligence conferences in the world, Gov. Gavin Newsom listened to two leaders in the field debate opposite views of a high-profile bill on his desk to protect Californians from the technology. "Honestly, I take advantage of opportunities like this," Newsom said recounting the exchange later during an interview at the Salesforce conference in San Francisco in mid-September. "I just watched them, and I was like, 'Here we go. Should I sign it, or should I not?' Then'absolutely,' 'absolutely not' and back and forth." The scene offered a peek into Newsom's deliberations on regulating the tech industry, including an explosion of AI companies, and the forces seeking to influence him during bill-signing season at the state Capitol.

Social dynamics in close human interactions pose significant challenges for Human Mesh Estimation (HME), particularly due to the complexity of physical contacts and the scarcity of training data. Addressing these challenges, we introduce a novel data generation method that utilizes Large Vision Language Models (LVLMs) to annotate contact maps which guide test-time optimization to produce paired image and pseudo-ground truth meshes. This methodology not only alleviates the annotation burden but also enables the assembly of a comprehensive dataset specifically tailored for close interactions in HME. Our Ask Pose Unite (APU) dataset, comprising over 6.2k human mesh pairs in contact covering diverse interaction types, is curated from images depicting naturalistic person-to-person scenes. We empirically show that using our dataset to train a diffusion-based contact prior, used as guidance during optimization, improves mesh estimation on unseen interactions. Our work addresses longstanding challenges of data scarcity for close interactions in HME enhancing the field's capabilities of handling complex interaction scenarios.

This research investigates the classification of Environmental, Social, and Governance (ESG) information within textual disclosures. The aim is to develop and evaluate binary classification models capable of accurately identifying and categorizing E, S and G-related content respectively. The motivation for this research stems from the growing importance of ESG considerations in investment decisions and corporate accountability. Accurate and efficient classification of ESG information is crucial for stakeholders to understand the impact of companies on sustainability and to make informed decisions. The research uses a quantitative approach involving data collection, data preprocessing, and the development of ESG-focused Large Language Models (LLMs) and traditional machine learning (Support Vector Machines, XGBoost) classifiers. Performance evaluation guides iterative refinement until satisfactory metrics are achieved. The research compares traditional machine learning techniques (Support Vector Machines, XGBoost), state-of-the-art language model (FinBERT-ESG) and fine-tuned LLMs like Llama 2, by employing standard Natural Language Processing performance metrics such as accuracy, precision, recall, F1-score. A novel fine-tuning method, Qlora, is applied to LLMs, resulting in significant performance improvements across all ESG domains. The research also develops domain-specific fine-tuned models, such as EnvLlama 2-Qlora, SocLlama 2-Qlora, and GovLlama 2-Qlora, which demonstrate impressive results in ESG text classification.