--Large Language Models (LLMs) excel in text classification, but their complexity hinders interpretability, making it difficult to understand the reasoning behind their predictions. Explainable AI (XAI) methods like LIME and SHAP offer local explanations by identifying influential words, but they rely on computationally expensive perturbations. These methods typically generate thousands of perturbed sentences and perform inferences on each, incurring a substantial computational burden, especially with LLMs. T o address this, we propose P erturbation-free L ocal Ex planation (PLEX), a novel method that leverages the contextual embeddings extracted from the LLM and a "Siamese network" style neural network trained to align with feature importance scores. This one-off training eliminates the need for subsequent perturbations, enabling efficient explanations for any new sentence. We demonstrate PLEX's effectiveness on four different classification tasks (sentiment, fake news, fake COVID-19 news and depression), showing more than 92% agreement with LIME and SHAP . Our evaluation using a "stress test" reveals that PLEX accurately identifies influential words, leading to a similar decline in classification accuracy as observed with LIME and SHAP when these words are removed. Notably, in some cases, PLEX demonstrates superior performance in capturing the impact of key features. PLEX dramatically accelerates explanation, reducing time and computational overhead by two and four orders of magnitude, respectively. This work offers a promising solution for explainable LLM-based text classification. ARGE language models (LLMs) have significantly advanced text classification, achieving state-of-the-art results in tasks like emotion recognition, sentiment analysis, topic categorization, and spam detection [1]. Powered by transformer architectures with millions or billions of parameters, they effectively capture complex linguistic patterns. However, the very complexity that enables their high performance also renders their internal workings opaque and difficult to interpret.

Transformers [1] have lead to breakthroughs in training large-scale general representations for computer vision (CV) and natural language processing (NLP) [2], enabling zero-shot adaptation and fast finetuning [3]. At the same time, despite impressive progress, transformer-based representations haven't shown the same versatility for robotic manipulation. Some attribute this gap to the lack of suitable training data for robotics [3]. We argue instead that data relevant to training robotic manipulation models is copious but has important structure that most existing training methods ignore and fail to leverage. These insights lead us to propose a novel transformer-based architecture, called PLEX, that is capable of effective learning from realistically available robotic manipulation datasets. We observe that robotics-relevant data falls into three major categories: (1) Video-only data, which contain high-quality and potentially description-annotated demonstrations for an immense variety of tasks but have no explicit action information for a robot to mimic; (2) Data containing matching sequences of percepts and actions, which are less plentiful than pure videos and don't necessarily correspond to meaningful tasks [4], but capture valuable correlations between a robot's actions and changes in the environment and are easy to collect on a given robot; (3) Small sets of high-quality sensorimotor demonstrations for a target task in a target environment. Thus, a scalable model architecture for robotic manipulation must be able to learn primarily from videos, while being extra data-efficient on sensorimotor training sequences and the small amount target demonstrations. PLEX, the PLanning-EXecution architecture we propose, is designed to take advantage of data sources of these types.

Plex now has the capability to skip intros and credits, so you don't have to sit through them if you don't want to. The streaming media service has introduced its new "Skip Credits" feature, a couple of years after a similar feature debuted for intros, and it shows up as a button at the bottom right corner of the screen. It appears the moment credits start rolling, allowing you to play the next episode in a series or to automatically jump to a mid- or post-credit scene in a movie with a single click. This new credit detection feature is available throughout Plex's entire free streaming catalogue, giving it a viewing experience similar to Netflix's in that regard. But you can also enable it for your personal collection, so long as you have Plex Pass subscription, which will set you back $5 a month or $40 a year. The company said it developed its own Skip Credits technology by training a machine learning algorithm to be able to detect markers like text and dark frames.

A recent trend in artificial intelligence is the use of pretrained models for language and vision tasks, which have achieved extraordinary performance but also puzzling failures. Probing these models' abilities in diverse ways is therefore critical to the field. In this paper, we explore the reliability of models, where we define a reliable model as one that not only achieves strong predictive performance but also performs well consistently over many decision-making tasks involving uncertainty (e.g., selective prediction, open set recognition), robust generalization (e.g., accuracy and proper scoring rules such as log-likelihood on in- and out-of-distribution datasets), and adaptation (e.g., active learning, few-shot uncertainty). We devise 10 types of tasks over 40 datasets in order to evaluate different aspects of reliability on both vision and language domains. To improve reliability, we developed ViT-Plex and T5-Plex, pretrained large model extensions for vision and language modalities, respectively. Plex greatly improves the state-of-the-art across reliability tasks, and simplifies the traditional protocol as it improves the out-of-the-box performance and does not require designing scores or tuning the model for each task. We demonstrate scaling effects over model sizes up to 1B parameters and pretraining dataset sizes up to 4B examples. We also demonstrate Plex's capabilities on challenging tasks including zero-shot open set recognition, active learning, and uncertainty in conversational language understanding.

Google this week unveiled a major redesign of Google Pay for Android and iOS. The app is meant to take on not just Apple Pay and Samsung Pay, but also PayPal, Venmo, and Mint all in one. Google also announced partnerships with 11 U.S. banks and credit unions to launch a mobile-first bank account service called Plex next year. Just like with health care, tech companies are becoming increasingly interested in banking. More competition in an industry that still hasn't embraced the internet, let alone the latest tech like artificial intelligence, is exciting.

Google Play Music is currently the best streaming music service for people who have their own music collections. The service lets users upload 50,000 of their own music files, then access the audio on a wide range of streaming devices. It's a great way to access your own music files from anywhere, and it doesn't cost a dime. Unfortunately, the free ride is just about over. At the end of this year, Google will discontinue Google Play Music and push users over to YouTube Music as a replacement.

This week's giveaway, courtesy of Plex, could put a sleek multimedia machine in your home. Plex's popular grid-view software helps you organize and navigate your files, while giving you access to them on-the-go. The NVIDIA Shield is a cord-cutting Android TV device with 4K HDR output, Google Assistant and GeForce Now game streaming on board. When paired with a Plex Pass, antenna and tuner you can harness OTA live TV with DVR capabilities, putting those shows seamlessly alongside your own library for easy access. Plex lets you access all this content through its app from a computer, mobile device and even supported systems like Sonos or Amazon's Alexa, wherever you go.

This undated image provided by Amazon.com, Inc. shows an Amazon Fire TV Cube. This year could be the year of the streaming service. Netflix, Amazon Prime, and Hulu have comfortably held the top three spots for streaming services, but this year they're going to see some serious brands looking to take a bite out of the streaming pie. IMDb, AT&T, Disney, Apple, and Warner have all announced or hinted at plans to launch streaming services in 2019.

This undated image provided by Amazon.com, Inc. shows an Amazon Fire TV Cube. Watch whatever you want, whenever you want. All you have to buy is one little box, and the monthly subscriptions are up to you. After decades of flipping through TV channels, many of us find the promise of internet-based television too miraculous to pass up.

Amazon's Alexa is becoming less hospitable to people who prefer to buy music instead of just signing up for a music subscription service such as Spotify. Until recently, Amazon Music Storage was the best way to stream your personal music collection onto Alexa devices--provided it was encoded as MP3 files. But in December, Amazon stopped supporting new uploads for the free service; it stopped accepting paid subscriptions a month later. In January 2019, Amazon will shut down Music Storage entirely, rendering your MP3 collection inaccessible on Echo devices unless you purchased the songs directly through Amazon's digital store. To play your own MP3s (or music encoded in other formats, such as FLAC) on the Echo or other Alexa devices, you can still use Plex or My Media Server for Alexa, both of which allow you to stream songs that you've stored on another device.