Large Language Models (LLMs) are driving advancements in artificial intelligence by increasing the scale of model parameters, which has significantly enhanced generalization ability and unlocked new capabilities in practice. However, their performance in specific downstream tasks is usually hindered by their knowledge boundaries on these tasks. Thus, fine-tuning techniques, especially the widely used Low-Rank Adaptation (LoRA) method, have been introduced to expand the boundaries on these tasks, whereas LoRA would underperform on certain tasks owing to its potential overfitting on these tasks. To overcome this overfitting and improve the performance of LoRA, we propose the flexible low rank adaptation (Flexora) method to automatically and flexibly select the most important layers needing to be fine-tuned to achieve the best performance on different downstream tasks. Specifically, Flexora firstly frames this layer selection problem as a well-defined hyperparameter optimization (HPO) problem, then addresses it using the unrolled differentiation (UD) method, and finally selects the most useful layers based on the optimized hyperparameters. Our extensive experiments on many pretrained models and natural language tasks show that Flexora is able to consistently improve over the existing baselines, indicating the effectiveness of our Flexora in practice. We additionally provide insightful theoretical results and many ablation studies to deliver a comprehensive understanding of our Flexora.

Adjusting robot behavior to human preferences can require intensive human feedback, preventing quick adaptation to new users and changing circumstances. Moreover, current approaches typically treat user preferences as a reward, which requires a manual balance between task success and user satisfaction. To integrate new user preferences in a zero-shot manner, our proposed Text2Interaction framework invokes large language models to generate a task plan, motion preferences as Python code, and parameters of a safe controller. By maximizing the combined probability of task completion and user satisfaction instead of a weighted sum of rewards, we can reliably find plans that fulfill both requirements. We find that 83% of users working with Text2Interaction agree that it integrates their preferences into the robot's plan, and 94% prefer Text2Interaction over the baseline. Our ablation study shows that Text2Interaction aligns better with unseen preferences than other baselines while maintaining a high success rate.

In Business Process Management (BPM), effectively comprehending process models is crucial yet poses significant challenges, particularly as organizations scale and processes become more complex. This paper introduces a novel framework utilizing the advanced capabilities of Large Language Models (LLMs) to enhance the interpretability of complex process models. We present different methods for abstracting business process models into a format accessible to LLMs, and we implement advanced prompting strategies specifically designed to optimize LLM performance within our framework. Additionally, we present a tool, AIPA, that implements our proposed framework and allows for conversational process querying. We evaluate our framework and tool by i) an automatic evaluation comparing different LLMs, model abstractions, and prompting strategies and ii) a user study designed to assess AIPA's effectiveness comprehensively. Results demonstrate our framework's ability to improve the accessibility and interpretability of process models, pioneering new pathways for integrating AI technologies into the BPM field.

The emergence of abundant electricity time series (ETS) data provides ample opportunities for various applications in the power systems, including demand-side management, grid stability, and consumer behavior analysis. Deep learning models have advanced ETS modeling by effectively capturing sequence dependence. Nevertheless, learning a generic representation of ETS data for various applications remains challenging due to the inherently complex hierarchical structure of ETS data. Moreover, ETS data exhibits intricate temporal dependencies and is suscepti ble to the influence of exogenous variables. Furthermore, different instances exhibit diverse electricity consumption behavior. In this paper, we propose a foundation model PowerPM to model ETS data, providing a large-scale, off-the-shelf model for power systems. PowerPM consists of a temporal encoder and a hierarchical encoder. The temporal encoder captures both temporal dependencies in ETS data, considering exogenous variables. The hierarchical encoder models the correlation between hierarchy. Furthermore, PowerPM leverages a novel self-supervised pretraining framework consisting of masked ETS modeling and dual-view contrastive learning, which enable PowerPM to capture temporal dependency within ETS windows and aware the discrepancy across ETS windows, providing two different perspectives to learn generic representation. Our experiments involve five real world scenario datasets, comprising private and public data. Through pre-training on massive ETS data, PowerPM achieves SOTA performance on diverse downstream tasks within the private dataset. Impressively, when transferred to the public datasets, PowerPM maintains its superiority, showcasing its remarkable generalization ability across various tasks and domains. Moreover, ablation studies, few-shot experiments provide additional evidence of the effectiveness of our model.

With the help of in-context learning (ICL), large language models (LLMs) have achieved impressive performance across various tasks. However, the function of descriptive instructions during ICL remains under-explored. In this work, we propose an ensemble prompt framework to describe the selection criteria of multiple in-context examples, and preliminary experiments on machine translation (MT) across six translation directions confirm that this framework boosts ICL perfromance. But to our surprise, LLMs might not necessarily care what the descriptions actually say, and the performance gain is primarily caused by the ensemble format, since the framework could lead to improvement even with random descriptive nouns. We further apply this new ensemble prompt on a range of commonsense, math, logical reasoning and hallucination tasks with three LLMs and achieve promising results, suggesting again that designing a proper prompt format would be much more effective and efficient than paying effort into specific descriptions. Our code will be publicly available once this paper is published.

In section 4, we discuss the consequences of to this question, we first argue that neither this for the current field that is structured around humans nor LLMs are general learners, benchmarks mostly concerned with measures of the in a variety of senses. We make a novel final, trained states of LLMs. In section 5, we apply case for how in particular LLMs follow a our arguments to the evaluations more focused dual-optimization process: they are optimized on the learning stages of LLMs. One debate asks during their training (which is typically whether LLMs are not too powerful, often phrases compared to language acquisition), around the question as to whether'impossible' languages, and modern LLMs have also been selected, that allegedly cannot be learned by humans, through a process akin to natural selection can be learned by LLMs. We add to the debate in a species. From this perspective, the fact that, even when trained to learn possible we argue that the performance of LLMs, languages, parts of the languages that LLMs whether similar or dissimilar to that of humans, learn are indeed impossible. This shows that the does not weigh easily on important biases of LLMs are different from ours, and remind debates about the importance of human us that an adequate model of learning has to learn cognitive biases for language.

Efforts directed towards promoting Open Government Data (OGD) have gained significant traction across various governmental tiers since the mid-2000s. As more datasets are published on OGD portals, finding specific data becomes harder, leading to information overload. Complete and accurate documentation of datasets, including association of proper tags with datasets is key to improving dataset findability and accessibility. Analysis conducted on the Estonian Open Data Portal, revealed that 11% datasets have no associated tags, while 26% had only one tag assigned to them, which underscores challenges in data findability and accessibility within the portal, which, according to the recent Open Data Maturity Report, is considered trend-setter. The aim of this study is to propose an automated solution to tagging datasets to improve data findability on OGD portals. This paper presents Tagify - a prototype of tagging interface that employs large language models (LLM) such as GPT-3.5-turbo and GPT-4 to automate dataset tagging, generating tags for datasets in English and Estonian, thereby augmenting metadata preparation by data publishers and improving data findability on OGD portals by data users. The developed solution was evaluated by users and their feedback was collected to define an agenda for future prototype improvements.

Condé Nast and OpenAI announced a multi-year partnership on Tuesday to display content from the publisher's brands such as the Vogue, Wired and the New Yorker within the AI startup's products, including ChatGPT and its SearchGPT prototype. The financial terms of the deal were not disclosed. The Microsoft-backed, Sam Altman-led firm has signed similar deals with Time magazine, the Financial Times, Business Insider owner Axel Springer, France's Le Monde and Spain's Prisa Media over the past few months. The deals give OpenAI access to the large archives of text owned by the publishers, which are necessary both for training large language models like ChatGPT and for finding real-time information. OpenAI launched its AI-powered search engine SearchGPT in July, with real-time access to information from the internet, making an incursion on territory long dominated by Google.

Condé Nast, the media conglomerate that owns publications like The New Yorker, Vogue and Wired, has announced a multi-year partnership OpenAI to display content from Condé Nast titles in ChatGPT as well as SearchGPT, the company's prototype AI-powered search engine. The partnership comes amid growing concerns over the unauthorized use of publishers' content by AI companies. Last month, Condé Nast sent a cease-and-desist letter to AI search startup Perplexity, accusing it of plagiarism for using its content to generate answers. "Over the last decade, news and digital media have faced steep challenges as many technology companies eroded publishers' ability to monetize content, most recently with traditional search," Condé Nast CEO Roger Lynch wrote to employees in a memo that was first reported by Semafor's Max Tani. "Our partnership with OpenAI begins to make up for some of that revenue, allowing us to continue to protect and invest in our journalism and creative endeavors."

Condé Nast and OpenAI have struck a multi-year deal that will allow the AI giant to use content from the media giant's roster of properties--which includes the New Yorker, Vogue, Vanity Fair, Bon Appetit, and, yes, WIRED. The deal will allow OpenAI to surface stories from these outlets in both ChatGPT and the new SearchGPT prototype. "It's crucial that we meet audiences where they are and embrace new technologies while also ensuring proper attribution and compensation for use of our intellectual property," Condé Nast CEO Roger Lynch wrote in a company-wide email. Lynch pointed to ongoing turmoil within the publishing industry while discussing the deal, noting that technology companies have made it harder for publishers to make money, most recently with changes to traditional search. "Our partnership with OpenAI begins to make up for some of that revenue, allowing us to continue to protect and invest in our journalism and creative endeavors," he wrote.