This paper highlights the importance of personalization in large language models and introduces the LaMP benchmark -- a novel benchmark for training and evaluating language models for producing personalized outputs. LaMP offers a comprehensive evaluation framework with diverse language tasks and multiple entries for each user profile. It consists of seven personalized tasks, spanning three text classification and four text generation tasks. We additionally propose two retrieval augmentation approaches that retrieve personal items from each user profile for personalizing language model outputs. To this aim, we study various retrieval models, including term matching, semantic matching, and time-aware methods. Extensive experiments on LaMP for zero-shot and fine-tuned language models demonstrate the efficacy of the proposed retrieval augmentation approach and highlight the impact of personalization in various natural language tasks.

In April 2022, when Dall-E, a text-to-image visio-linguistic model, was released, it purportedly attracted over a million users within the first three months. This was followed by ChatGPT, in January 2023, which apparently reached 100 million monthly active users just two months after launch. Both mark notable moments in the development of generative AI, which in turn has brought forth an explosion of AI-generated content into the web. The bad news is that, in 2024, this means we will also see an explosion of fabricated, nonsensical information, mis- and disinformation, and the exacerbation of social negative stereotypes encoded in these AI models. The AI revolution wasn't spurred by any recent theoretical breakthrough--indeed, most of the foundational work underlying artificial neural networks has been around for decades--but by the "availability" of massive data sets.

The piece seems to conjure a prehistoric avant-garde musical workshop, a sonic analogue of the visual culture that can be glimpsed in the cave. Fully notated passages--scampering runs, precisely hammering chords, ghostly arpeggios--are interspersed with opportunities for improvisation. The first twenty-four bars indicate rhythms, dynamics, and registers but not precise pitches. The ending, too, is left open. Cory Smythe, himself a composer and improviser of note, proved an ideal conduit, making the distinction between Lewis's ideas and his own elaborations inconsequential.

There was a time when smartphone makers rushed to quad-camera claims, most of which did so by throwing in a mediocre fourth camera -- usually for macro shots, if not a monochrome filter or just a depth sensor. Nowadays, though, we are blessed with legit quad "main" cameras on some flagship phones, so brands need to be more creative to further differentiate themselves. In Oppo's case, it decided to feature not just one, but two periscope telephoto cameras on its new Snapdragon 8 Gen 3-powered Find X7 Ultra, as a leap from the triple camera system on the previous model. The Find X7 Ultra's "HyperTone Camera System" features the same 50-megapixel resolution across all four rear Hasselblad cameras, thanks to their relatively large sensors compared to the competition, according to Oppo. The main imager packs Sony's second-gen 1-inch sensor, the LYT-900, which is more efficient in terms of power consumption and thermal performance. This is complemented by an f/1.8 aperture, OIS (optical image stabilization), a 23mm focal length and a 50-percent reduction in lens reflection.

As large language models continue to advance in Artificial Intelligence (AI), text generation systems have been shown to suffer from a problematic phenomenon termed often as "hallucination." However, with AI's increasing presence across various domains including medicine, concerns have arisen regarding the use of the term itself. In this study, we conducted a systematic review to identify papers defining "AI hallucination" across fourteen databases. We present and analyze definitions obtained across all databases, categorize them based on their applications, and extract key points within each category. Our results highlight a lack of consistency in how the term is used, but also help identify several alternative terms in the literature. We discuss implications of these and call for a more unified effort to bring consistency to an important contemporary AI issue that can affect multiple domains significantly.

In recent years, various well-designed algorithms have empowered music platforms to provide content based on one's preferences. Music genres are defined through various aspects, including acoustic features and cultural considerations. Music genre classification works well with content-based filtering, which recommends content based on music similarity to users. Given a considerable dataset, one premise is automatic annotation using machine learning or deep learning methods that can effectively classify audio files. The effectiveness of systems largely depends on feature and model selection, as different architectures and features can facilitate each other and yield different results. In this study, we conduct a comparative study investigating the performances of three models: a proposed convolutional neural network (CNN), the VGG16 with fully connected layers (FC), and an eXtreme Gradient Boosting (XGBoost) approach on different features: 30-second Mel spectrogram and 3-second Mel-frequency cepstral coefficients (MFCCs). The results show that the MFCC XGBoost model outperformed the others. Furthermore, applying data segmentation in the data preprocessing phase can significantly enhance the performance of the CNNs.

Automatically understanding funny moments (i.e., the moments that make people laugh) when watching comedy is challenging, as they relate to various features, such as body language, dialogues and culture. In this paper, we propose FunnyNet-W, a model that relies on cross- and self-attention for visual, audio and text data to predict funny moments in videos. Unlike most methods that rely on ground truth data in the form of subtitles, in this work we exploit modalities that come naturally with videos: (a) video frames as they contain visual information indispensable for scene understanding, (b) audio as it contains higher-level cues associated with funny moments, such as intonation, pitch and pauses and (c) text automatically extracted with a speech-to-text model as it can provide rich information when processed by a Large Language Model. To acquire labels for training, we propose an unsupervised approach that spots and labels funny audio moments. We provide experiments on five datasets: the sitcoms TBBT, MHD, MUStARD, Friends, and the TED talk UR-Funny. Extensive experiments and analysis show that FunnyNet-W successfully exploits visual, auditory and textual cues to identify funny moments, while our findings reveal FunnyNet-W's ability to predict funny moments in the wild. FunnyNet-W sets the new state of the art for funny moment detection with multimodal cues on all datasets with and without using ground truth information.

In recent years, DNNs have shown promising results This paper describes methods for training autonomous in the field of behavioural cloning (BC) [5, 18]. BC is a agents to play the game "Doom 2" through Imitation form of Imitation Learning (IL), where we train an artificial Learning (IL) using only pixel data as input. We also explore "agent" to mimic actions from an observable state of how Reinforcement Learning (RL) compares to IL expert data [34]. Agents are trained using a number of historical for humanness by comparing camera movement and trajectory states, be they image frames or other data, and their data. Through behavioural cloning, we examine the corresponding actions. The learning is performed by using ability of individual models to learn varying behavioural the final frame's associated action as the "target", this target traits. We attempt to mimic the behaviour of real players being passed to some loss function. The loss function will with different play styles, and find we can train agents that reinforce the observed frame's predicted action, doing this behave aggressively, passively, or simply more human-like over an extremely large dataset will achieve an agent that than traditional AIs. We propose these methods of introducing can predict the best action to take at any one given set of more depth and human-like behaviour to agents in video input image frames [17].

Large language models with instruction-following abilities have revolutionized the field of artificial intelligence. These models show exceptional generalizability to tackle various real-world tasks through their natural language interfaces. However, their performance heavily relies on high-quality exemplar data, which is often difficult to obtain. This challenge is further exacerbated when it comes to multimodal instruction following. We introduce TextBind, an almost annotation-free framework for empowering larger language models with the multi-turn interleaved multimodal instruction-following capabilities. Our approach requires only image-caption pairs and generates multi-turn multimodal instruction-response conversations from a language model. To accommodate interleaved image-text inputs and outputs, we devise MIM, a language model-centric architecture that seamlessly integrates image encoder and decoder models. We release our dataset, model, and demo to foster future research in the area of multimodal instruction following.

Instruction tuning has emerged as a promising approach to enhancing large language models in following human instructions. It is shown that increasing the diversity and number of instructions in the training data can consistently enhance generalization performance, which facilitates a recent endeavor to collect various instructions and integrate existing instruction tuning datasets into larger collections. However, different users have their unique ways of expressing instructions, and there often exist variations across different datasets in the instruction styles and formats, i.e., format inconsistency. In this work, we propose a framework named Unified Instruction Tuning (UIT), which calls OpenAI APIs for automatic format transfer among different instruction tuning datasets such as PromptSource, FLAN and CrossFit. With the framework, we (1) demonstrate the necessity of maintaining format consistency in instruction tuning; (2) improve the generalization performance on unseen instructions on T5-LM-xl; (3) provide a novel perplexity-based denoising method to reduce the noise of automatic format transfer to make the UIT framework more practical and a smaller offline model based on GPT-J that achieves comparable format transfer capability to OpenAI APIs to reduce costs in practice. Further analysis regarding variations of targeted formats and other effects is intended.