With the growth of social media and large language models, content moderation has become crucial. Many existing datasets lack adequate representation of different groups, resulting in unreliable assessments. To tackle this, we propose a socio-culturally aware evaluation framework for LLM-driven content moderation and introduce a scalable method for creating diverse datasets using persona-based generation. Our analysis reveals that these datasets provide broader perspectives and pose greater challenges for LLMs than diversity-focused generation methods without personas. This challenge is especially pronounced in smaller LLMs, emphasizing the difficulties they encounter in moderating such diverse content.

This paper introduces a new approach to generating strongly constrained texts. We consider standardized sentence generation for the typical application of vision screening. To solve this problem, we formalize it as a discrete combinatorial optimization problem and utilize multivalued decision diagrams (MDD), a well-known data structure to deal with constraints. In our context, one key strength of MDD is to compute an exhaustive set of solutions without performing any search. Once the sentences are obtained, we apply a language model (GPT-2) to keep the best ones. We detail this for English and also for French where the agreement and conjugation rules are known to be more complex. Finally, with the help of GPT-2, we get hundreds of bona-fide candidate sentences. When compared with the few dozen sentences usually available in the well-known vision screening test (MNREAD), this brings a major breakthrough in the field of standardized sentence generation. Also, as it can be easily adapted for other languages, it has the potential to make the MNREAD test even more valuable and usable. More generally, this paper highlights MDD as a convincing alternative for constrained text generation, especially when the constraints are hard to satisfy, but also for many other prospects.

This work aims to employ natural language generation (NLG) to rapidly generate items for English language learning applications: this requires both language models capable of generating fluent, high-quality English, and to control the output of the generation to match the requirements of the relevant items. We experiment with deep pretrained models for this task, developing novel methods for controlling items for factors relevant in language learning: diverse sentences for different proficiency levels and argument structure to test grammar. Human evaluation demonstrates high grammatically scores for all models (3.4 and above out of 4), and higher length (24%) and complexity (9%) over the baseline for the advanced proficiency model. Our results show that we can achieve strong performance while adding additional control to ensure diverse, tailored content for individual users.

This paper tackles recipe generation from unsegmented cooking videos, a task that requires agents to (1) extract key events in completing the dish and (2) generate sentences for the extracted events. Our task is similar to dense video captioning (DVC), which aims at detecting events thoroughly and generating sentences for them. However, unlike DVC, in recipe generation, recipe story awareness is crucial, and a model should output an appropriate number of key events in the correct order. We analyze the output of the DVC model and observe that although (1) several events are adoptable as a recipe story, (2) the generated sentences for such events are not grounded in the visual content. Based on this, we hypothesize that we can obtain correct recipes by selecting oracle events from the output events of the DVC model and re-generating sentences for them. To achieve this, we propose a novel transformer-based joint approach of training event selector and sentence generator for selecting oracle events from the outputs of the DVC model and generating grounded sentences for the events, respectively. In addition, we extend the model by including ingredients to generate more accurate recipes. The experimental results show that the proposed method outperforms state-of-the-art DVC models. We also confirm that, by modeling the recipe in a story-aware manner, the proposed model output the appropriate number of events in the correct order.

A novel approach to automated learning of syntactic rules governing natural languages is proposed, based on using probabilities assigned to sentences (and potentially longer word sequences) by transformer neural network language models to guide symbolic learning processes like clustering and rule induction. This method exploits the learned linguistic knowledge in transformers, without any reference to their inner representations; hence, the technique is readily adaptable to the continuous appearance of more powerful language models. We show a proof-of-concept example of our proposed technique, using it to guide unsupervised symbolic link-grammar induction methods drawn from our prior research.

Text style transfer aims to modify the style of a sentence while keeping its content unchanged. Recent style transfer systems often fail to faithfully preserve the content after changing the style. In particular, we achieve the goal by devising rich model objectives based on both the sentence's lexical information and a language model that conditions on content. The resulting model therefore is encouraged to retain the semantic meaning of the target sentences. We perform extensive experiments that compare our model to other existing approaches in the tasks of sentiment and political slant transfer. Our model achieves significant improvement in terms of both content preservation and style transfer in automatic and human evaluation. Text style transfer is an important task in designing sophisticated and controllable natural language generation (NLG) systems. The goal of this task is to convert a sentence from one style (e.g., negative sentiment) to another (e.g., positive sentiment), while preserving the style-independent content (e.g., the name of the food being discussed). Typically, it is difficult to find parallel data with different styles. So we must learn to disentangle the representations of the style from the content. However, it is impossible to separate the two components by simply adding or dropping certain words.

By now, you probably already know that artificial intelligence (AI) technology is developing at a remarkably fast rate. It has been the subject of many essays and news articles that are mainly about an oncoming robot takeover, which could mean trouble for many humans in the world today. But in order for advanced AI robots to truly take over the world, they've got to first be able to think like humans. Now a group of researchers have taken the first steps to making that a reality. Japanese researchers recently published a new study titled, "Describing Semantic Representations of Brain Activity Evoked by Visual Stimuli," wherein AI technology was used to predict – with great accuracy – exactly what people were thinking when they were looking at certain pictures.

Think that Google's search algorithms are good at reading your mind? That's nothing compared to a new artificial intelligence research project coming out of Japan, which can analyze a person's brain scans and provide a written description of what they have been looking at. To generate its captions, the artificial intelligence is given an fMRI brain scan image, taken while a person is looking at a picture. It then generates a written description of what they think the person was viewing. An illustration of the level of complexity it can offer is: "A dog is sitting on the floor in front of an open door" or "a group of people standing on the beach."

Think that Google's search algorithms are good at reading your mind? That's nothing compared to a new artificial intelligence research project coming out of Japan, which can analyze a person's brain scans and provide a written description of what they have been looking at. To generate its captions, the artificial intelligence is given an fMRI brain scan image, taken while a person is looking at a picture. It then generates a written description of what they think the person was viewing. An illustration of the level of complexity it can offer is: "A dog is sitting on the floor in front of an open door" or "a group of people standing on the beach."

Generating plausible and fluent sentence with desired properties has long been a challenge. Most of the recent works use recurrent neural networks (RNNs) and their variants to predict following words given previous sequence and target label. In this paper, we propose a novel framework to generate constrained sentences via Gibbs Sampling. The candidate sentences are revised and updated iteratively, with sampled new words replacing old ones. Our experiments show the effectiveness of the proposed method to generate plausible and diverse sentences.