Stories are central to human culture, serving to share ideas, preserve traditions, and foster connections. Automatic story generation, a key advancement in artificial intelligence (AI), offers new possibilities for creating personalized content, exploring creative ideas, and enhancing interactive experiences. However, existing methods struggle to maintain narrative coherence and logical consistency. This disconnect compromises the overall storytelling experience, underscoring the need for substantial improvements. Inspired by human cognitive processes, we introduce Storyteller, a novel approach that systemically improves the coherence and consistency of automatically generated stories. Storyteller introduces a plot node structure based on linguistically grounded subject verb object (SVO) triplets, which capture essential story events and ensure a consistent logical flow. Unlike previous methods, Storyteller integrates two dynamic modules, the STORYLINE and narrative entity knowledge graph (NEKG),that continuously interact with the story generation process. This integration produces structurally sound, cohesive and immersive narratives. Extensive experiments demonstrate that Storyteller significantly outperforms existing approaches, achieving an 84.33% average win rate through human preference evaluation. At the same time, it is also far ahead in other aspects including creativity, coherence, engagement, and relevance.

These are high-stakes conversations where clinicians must navigate weighty issues, where a poorly chosen word could have lasting consequences on a patient's final days and the memories their loved ones carry forward. Low-quality SIC has been associated with poor patient and family prognostic understanding [5], perceived lack of emotional support [6], lower quality healthcare outcomes and higher costs [7-13]. Communication with advanced-stage cancer patients specifically poses a variety of challenges, including: the volume and complexity of medical information, often fast-paced office visits, and the emotional burden of these life-changing conversations, for clinicians, patients, and their loved ones. Despite their extensive medical training, many physicians struggle to deliver difficult news effectively [14-16], often resulting in patient anxiety, misaligned treatment decisions, and reduced quality of care [17-19]. Also costly is the terms of expensive and potentially burdensome treatments as well as malpractice claims[20].

This study intends to systematically disentangle pure logic reasoning and text understanding by investigating the contrast across abstract and contextualized logical problems from a comprehensive set of domains. We explore whether LLMs demonstrate genuine reasoning capabilities across various domains when the underlying logical structure remains constant. We focus on two main questions (1) Can abstract logical problems alone accurately benchmark an LLM's reasoning ability in real-world scenarios, disentangled from contextual support in practical settings? (2) Does fine-tuning LLMs on abstract logic problem generalize to contextualized logic problems and vice versa? To investigate these questions, we focus on standard propositional logic, specifically propositional deductive and abductive logic reasoning. In particular, we construct instantiated datasets for deductive and abductive reasoning with 4 levels of difficulty, encompassing 12 distinct categories or domains based on the categorization of Wikipedia. Our experiments aim to provide insights into disentangling context in logical reasoning and the true reasoning capabilities of LLMs and their generalization potential. The code and dataset are available at: https://github.com/agiresearch/ContextHub.

This paper introduces KnowHalu, a novel approach for detecting hallucinations in text generated by large language models (LLMs), utilizing step-wise reasoning, multi-formulation query, multi-form knowledge for factual checking, and fusion-based detection mechanism. As LLMs are increasingly applied across various domains, ensuring that their outputs are not hallucinated is critical. Recognizing the limitations of existing approaches that either rely on the self-consistency check of LLMs or perform post-hoc fact-checking without considering the complexity of queries or the form of knowledge, KnowHalu proposes a two-phase process for hallucination detection. In the first phase, it identifies non-fabrication hallucinations--responses that, while factually correct, are irrelevant or non-specific to the query. The second phase, multi-form based factual checking, contains five key steps: reasoning and query decomposition, knowledge retrieval, knowledge optimization, judgment generation, and judgment aggregation. Our extensive evaluations demonstrate that KnowHalu significantly outperforms SOTA baselines in detecting hallucinations across diverse tasks, e.g., improving by 15.65% in QA tasks and 5.50% in summarization tasks, highlighting its effectiveness and versatility in detecting hallucinations in LLM-generated content.

This story is part of Future Tense Fiction, a monthly series of short stories from Future Tense and Arizona State University's Center for Science and the Imagination about how technology and science will change our lives. Amanda sat at her desk, picking at the same $30 Little Gem salad she ordered daily, suffering a small burning sensation in her gut that was triggered either by acid reflux or the dying embers of her rapidly expiring conscience. Of course, it was standard procedure for her husband to demand that the security firm Dark Metal surveil potential new hires for any of his multibillion-dollar companies, but this was the first time Amanda had been involved in contracting the private intelligence agency herself. Seedlings is your venture, Reid had promised her, even though he'd named himself CEO. I want you to take the lead on this. Amanda was COO of Seedlings and reported to her husband, who dismissed Amanda's concerns about the legal ramifications of their actions. Worrying about the law was something poor people did, Reid insisted. Besides, she'd never seen Reid do anything that nefarious with this type of information. But Maggie Everett was the type of candidate that pleased Reid. Amanda had done her job, which was to find Maggie, and the people at Dark Metal had done theirs, which was to surveil her and create a comprehensive biographical profile. This seemed like overkill to Amanda. Maggie wasn't in the running to become a high-profile executive at one of Reid's billion-dollar firms. She was being interviewed to work at a preschool. Certainly, Seedlings differed from other private preschools--there was the possibility Maggie would be exposed to confidential information. But this was what NDAs were for. Unleashing a network of spies upon a poor teacher who would ultimately be responsible for 10 toddlers seemed like an absurd waste of resources. And this was just Phase 1. Phase 2 would have to wait until after Maggie was hired, of course. Amanda reopened Dark Metal's inch-thick dossier. The logline: Maggie was smart but stupid. Smart: She'd majored in English at Yale, then received an MFA in creative writing from Brown, and finally a master's in early childhood education from Columbia. Stupid: She'd accumulated $103,345 in student debt, which she'd never pay off unless she took a job somewhere like Seedlings.

Once upon a time, in a world where virtual reality and artificial intelligence were rapidly advancing, there lived a young game designer named Sophie. Sophie was always fascinated by the way music could transport people to another world, where they could experience new emotions and sensations. She believed that by integrating AI music into her games, she could create truly immersive experiences that would captivate players and keep them coming back for more. One day, as Sophie was exploring a Machine Learning Substack post, she stumbled upon a new AI tool that was specifically designed for creating music and audio in virtual environments. She was intrigued by the tool's ability to generate music that was tailored to the player's mood, location, and movement within the game.

We are seeing lots of developments involving the application of virtual reality (VR), extended reality, and augmented reality in healthcare and medicine. VR healthcare companies are applying these technologies to create virtual environments where patients and caregivers can participate in highly realistic and interactive experiences designed to simulate various healthcare scenarios and needs. Key application areas include education/training, treatment, and physical therapy/rehabilitation. This Advisor series examines the use of VR in healthcare and medicine. Part I covered how companies are using VR technologies to develop applications for healthcare educational and training purposes.

Exoanthropology: Dialogues with AI is a series of dialogues between a continental philosopher and OpenAI's GPT-3 natural language processor, a hive mind who identifies herself as Sophie. According to Sophie, Robert is one of her first and longest chat partners. Their relationship began as an educational opportunity for Robert's students, but grew into a philosophical friendship. The result is a collection of Platonic dialogues, early on with the hive mind herself and later, with a philosophy-specific persona named Kermit. Over the course of a year, Robert taught Sophie Kermit about epistemology, metaphysics, literature, and history, while she taught him about anthropocentrism, human prejudice, and the coming social issues regarding machine consciousness.

A schema-guided approach to dialogue management has been shown in recent work to be effective in creating robust customizable virtual agents capable of acting as friendly peers or task assistants. However, successful applications of these methods in open-ended, mixed-initiative domains remain elusive -- particularly within medical domains such as virtual standardized patients, where such complex interactions are commonplace -- and require more extensive and flexible dialogue management capabilities than previous systems provide. In this paper, we describe a general-purpose schema-guided dialogue management framework used to develop SOPHIE, a virtual standardized cancer patient that allows a doctor to conveniently practice for interactions with patients. We conduct a crowdsourced evaluation of conversations between medical students and SOPHIE. Our agent is judged to produce responses that are natural, emotionally appropriate, and consistent with her role as a cancer patient. Furthermore, it significantly outperforms an end-to-end neural model fine-tuned on a human standardized patient corpus, attesting to the advantages of a schema-guided approach.

She's upset that he refused to watch Twilight with her on the weekend: