Of all the egomaniacal lions who ruled Hollywood during the 20th century gatekeeper era, very few made a brilliant pivot to the internet. The exception is Barry Diller. After leading programming at ABC, running Paramount, and supercharging Fox by launching its broadcast network in the late 1980s, Diller no longer wanted to work for anyone else. Either you are or you aren't, he said of independence. As a free agent he quickly grasped the power of interactivity and built an empire that includes Expedia Group, almost the entire online dating sector (Tinder, Match, OkCupid), and an online media lineup that includes People, which wrote a hit piece on him early in his career titled "Failing Upwards."

What if you could turn your everyday snapshots into jaw-dropping works of art with just a few clicks? That's exactly what Luminar Neo does--and right now, you can get lifetime access to this award-winning photo-editing software for just 89.99 (regularly 682). Plus, you'll get a ton of exclusive add-ons that'll elevate your photography game from amateur to absolutely on point. It's AI-powered, intuitive, and packed with state-of-the-art tools that make even the most complex edits feel effortless. Whether you're a pro photographer, content creator, or just love tinkering with your travel pics, Luminar Neo's tools--like Sky for instant sky replacements and Relight to adjust lighting like a pro--make magic happen.

While at the park, the service members had the chance to explore attractions and participate in Disneyland's daily flag ceremony. Disney's Imagineers are working on a new animatronic of iconic American visionary Walt Disney, but some members of his family have opposing views about whether it celebrates his legacy or dehumanizes him. Disney's Main Street Opera House plans to unveil a new theme park attraction called Walt Disney – A Magical Life, featuring an audio-animatronic of the company's founder. But Joanna Miller, one of Disney's grandchildren, slammed the idea of an animatronic as "dehumanizing" in a viral Facebook post. Among her claims, she suggested that her grandfather had told early Imagineer Sam McKim he never wanted to be commemorated with an animatronic.

The progress of AI systems such as large language models (LLMs) raises increasingly pressing concerns about their safe deployment. This paper examines the value alignment problem for LLMs, arguing that current alignment strategies are fundamentally inadequate to prevent misuse. Despite ongoing efforts to instill norms such as helpfulness, honesty, and harmlessness in LLMs through fine-tuning based on human preferences, they remain vulnerable to adversarial attacks that exploit conflicts between these norms. I argue that this vulnerability reflects a fundamental limitation of existing alignment methods: they reinforce shallow behavioral dispositions rather than endowing LLMs with a genuine capacity for normative deliberation. Drawing from on research in moral psychology, I show how humans' ability to engage in deliberative reasoning enhances their resilience against similar adversarial tactics. LLMs, by contrast, lack a robust capacity to detect and rationally resolve normative conflicts, leaving them susceptible to manipulation; even recent advances in reasoning-focused LLMs have not addressed this vulnerability. This ``shallow alignment'' problem carries significant implications for AI safety and regulation, suggesting that current approaches are insufficient for mitigating potential harms posed by increasingly capable AI systems.

Automatic fact-checking has recently received more attention as a means of combating misinformation. Despite significant advancements, fact-checking systems based on retrieval-augmented language models still struggle to tackle adversarial claims, which are intentionally designed by humans to challenge fact-checking systems. To address these challenges, we propose a training-free method designed to rephrase the original claim, making it easier to locate supporting evidence. Our modular framework, SUCEA, decomposes the task into three steps: 1) Claim Segmentation and Decontextualization that segments adversarial claims into independent sub-claims; 2) Iterative Evidence Retrieval and Claim Editing that iteratively retrieves evidence and edits the subclaim based on the retrieved evidence; 3) Evidence Aggregation and Label Prediction that aggregates all retrieved evidence and predicts the entailment label. Experiments on two challenging fact-checking datasets demonstrate that our framework significantly improves on both retrieval and entailment label accuracy, outperforming four strong claim-decomposition-based baselines.

Misleading text detection on social media platforms is a critical research area, as these texts can lead to public misunderstanding, social panic and even economic losses. This paper proposes a novel framework - CL-ISR (Contrastive Learning and Implicit Stance Reasoning), which combines contrastive learning and implicit stance reasoning, to improve the detection accuracy of misleading texts on social media. First, we use the contrastive learning algorithm to improve the model's learning ability of semantic differences between truthful and misleading texts. Contrastive learning could help the model to better capture the distinguishing features between different categories by constructing positive and negative sample pairs. This approach enables the model to capture distinguishing features more effectively, particularly in linguistically complicated situations. Second, we introduce the implicit stance reasoning module, to explore the potential stance tendencies in the text and their relationships with related topics. This method is effective for identifying content that misleads through stance shifting or emotional manipulation, because it can capture the implicit information behind the text. Finally, we integrate these two algorithms together to form a new framework, CL-ISR, which leverages the discriminative power of contrastive learning and the interpretive depth of stance reasoning to significantly improve detection effect.

Existing studies of innovation emphasize the power of social structures to shape innovation capacity. Emerging machine learning approaches, however, enable us to model innovators' personal perspectives and interpersonal innovation opportunities as a function of their prior trajectories of experience. We theorize then quantify subjective perspectives and innovation opportunities based on innovator positions within the geometric space of concepts inscribed by dynamic language representations. Using data on millions of scientists, inventors, writers, entrepreneurs, and Wikipedia contributors across the creative domains of science, technology, film, entrepreneurship, and Wikipedia, here we show that measured subjective perspectives anticipate what ideas individuals and groups creatively attend to and successfully combine in future. When perspective and background diversity are decomposed as the angular difference between collaborators' perspectives on their creation and between their experiences, the former consistently anticipates creative achievement while the latter portends its opposite, across all cases and time periods examined. We analyze a natural experiment and simulate creative collaborations between AI (large language model) agents designed with various perspective and background diversity, which are consistent with our observational findings. We explore mechanisms underlying these findings and identify how successful collaborators leverage common language to weave together diverse experience obtained through trajectories of prior work that converge to provoke one another and innovate. We explore the importance of these findings for team assembly and research policy.

Learning from preference feedback is essential for aligning large language models (LLMs) with human values and improving the quality of generated responses. However, existing preference learning methods rely heavily on curated data from humans or advanced LLMs, which is costly and difficult to scale. In this work, we present PUGC, a novel framework that leverages implicit human Preferences in unlabeled User-Generated Content (UGC) to generate preference data. Although UGC is not explicitly created to guide LLMs in generating human-preferred responses, it often reflects valuable insights and implicit preferences from its creators that has the potential to address readers' questions. PUGC transforms UGC into user queries and generates responses from the policy model. The UGC is then leveraged as a reference text for response scoring, aligning the model with these implicit preferences. This approach improves the quality of preference data while enabling scalable, domain-specific alignment. Experimental results on Alpaca Eval 2 show that models trained with DPO and PUGC achieve a 9.37% performance improvement over traditional methods, setting a 35.93% state-of-the-art length-controlled win rate using Mistral-7B-Instruct. Further studies highlight gains in reward quality, domain-specific alignment effectiveness, robustness against UGC quality, and theory of mind capabilities. Our code and dataset are available at https://zhaoxuan.info/PUGC.github.io/

Red teaming has proven to be an effective method for identifying and mitigating vulnerabilities in Large Language Models (LLMs). Reinforcement Fine-Tuning (RFT) has emerged as a promising strategy among existing red teaming techniques. However, a lack of a unified benchmark hinders current RFT-based red teaming methods. Implementation details, especially in Proximal Policy Optimization (PPO)-based RFT, significantly affect outcome stability and reproducibility. To address this issue, we introduce RedRFT, a lightweight benchmark designed to simplify and standardize the implementation and evaluation of RFT-based red teaming. RedRFT combines the design strengths of both single-file CleanRL and highly modularized Tianshou, offering high-quality single-file red teaming implementations and modular PPO core components, such as the General Advantage Estimator. It supports a variety of token and sentence diversity metrics, featuring modularized intrinsic reward computation that facilitates plug-and-play experimentation. To clarify their influence on RFT performance, we conducted an extensive ablation study on key components, including Low-Rank Adaptation (LoRA), Kullback-Leibler (KL) divergence, and Lagrange Multiplier. We hope this work contributes to 1) gaining a comprehensive understanding of the implementation nuances of RFT-based red teaming algorithms, and 2) enabling rapid prototyping of innovative features for RFT-based red teaming. Code for the benchmark can be accessed at https://github.com/x-zheng16/RedRFT.git.

Fine-tuning large language models (LLMs) to adapt to evolving safety policies is costly and impractical. Mechanistic interpretability enables inference-time control through latent activation steering, yet its potential for precise, customizable safety adjustments remains largely untapped. This paper investigates an approach called SafeSteer for guiding the outputs of LLMs by: (i) leveraging category-specific steering vectors for more precise control, (ii) employing a simple, gradient-free unsupervised method to enhance safety steering while preserving text quality, topic relevance, and without explicit refusal, and (iii) accomplishing this without a hard requirement of contrastive pairwise safe data. We also highlight that our method, being simple and effective, aligns with recent studies suggesting that simple techniques often outperform more complex ones in activation steering. We showcase the effectiveness of our approach across various LLMs, datasets, and risk categories, demonstrating its ability to provide precise control, prevent blanket refusals, and guide models toward generating safe content while maintaining topic relevance.