With the rising popularity of Transformer-based large language models (LLMs), reducing their high inference costs has become a significant research focus. One effective approach is to compress the long input contexts....
详细信息
Recent developments show that Large language Models (LLMs) produce state-of-the-art performance on naturallanguage (NL) to code generation for resource-rich general-purpose languages like C++, Java, and ***, their pr...
详细信息
Planning for goal-oriented dialogue often requires simulating future dialogue interactions and estimating task progress. Many approaches thus consider training neural networks to perform look-ahead search algorithms s...
详细信息
ISBN:
(纸本)9798891760608
Planning for goal-oriented dialogue often requires simulating future dialogue interactions and estimating task progress. Many approaches thus consider training neural networks to perform look-ahead search algorithms such as A* search and Monte Carlo Tree Search (MCTS). However, this training often requires abundant annotated data, which creates challenges when faced with noisy annotations or low-resource settings. We introduce GDP- ZERO, an approach using Open-Loop MCTS to perform goal-oriented dialogue policy planning without any model training. GDP-ZERO prompts a large language model to act as a policy prior, value function, user simulator, and system model during the tree search. We evaluate GDP- ZERO on the goal-oriented task PersuasionForGood, and find that its responses are preferred over ChatGPT up to 59.32% of the time, and are rated more persuasive than Chat-GPT during interactive evaluations(1).
Many annotation tasks in naturallanguageprocessing are highly subjective in that there can be different valid and justified perspectives on what is a proper label for a given example. This also applies to the judgme...
详细信息
ISBN:
(纸本)9798891760608
Many annotation tasks in naturallanguageprocessing are highly subjective in that there can be different valid and justified perspectives on what is a proper label for a given example. This also applies to the judgment of argument quality, where the assignment of a single ground truth is often questionable. At the same time, there are generally accepted concepts behind argumentation that form a common ground. To best represent the interplay of individual and shared perspectives, we consider a continuum of approaches ranging from models that fully aggregate perspectives into a majority label to "share nothing"-architectures in which each annotator is considered in isolation from all other annotators. In between these extremes, inspired by models used in the field of recommender systems, we investigate the extent to which architectures that include layers to model the relations between different annotators are beneficial for predicting single-annotator labels. By means of two tasks of argument quality classification (argument concreteness and validity/novelty of conclusions), we show that recommender architectures increase the averaged annotator-individual F-1-scores up to 43% over a majority-label model. Our findings indicate that approaches to subjectivity can benefit from relating individual perspectives.
Generative language models often struggle with specialized or less-discussed knowledge. A potential solution is found in Retrieval-Augmented Generation (RAG) models which act like retrieving information before generat...
详细信息
As one of the common rhetorical devices, puns play a vital role in linguistic study, including the comprehensive analysis of linguistic humor. Although large language models (LLMs) have been widely explored on various...
详细信息
There is a growing interest in expanding the input capacity of language models (LMs) across various domains. However, simply increasing the context window does not guarantee robust performance across diverse long-inpu...
详细信息
We introduce GenerativeDictionary, a novel dictionary system that generates word sense interpretations based on the given context. Our approach involves transforming context sentences to highlight the meaning of targe...
详细信息
We study the presence of heteronormative biases and prejudice against interracial romantic relationships in large language models by performing controlled name-replacement experiments for the task of relationship pred...
详细信息
Owing to the increased video content consumption in recent years, the need for advanced contextual advertising methods that leverage increasing user engagement and relevance on advertisement-based video-on-demand plat...
详细信息
Owing to the increased video content consumption in recent years, the need for advanced contextual advertising methods that leverage increasing user engagement and relevance on advertisement-based video-on-demand platforms has increased. Traditional behavior-based advertisement targeting is waning, particularly owing to the recent strict privacy policies that favor user consent and privacy. This study proposes an innovative approach for integrating advanced naturallanguageprocessing with multimodal analysis for video contextual advertising. To this end, transformer-based architectures, specifically BERTopic, computer vision techniques, and large language models were used to extract sets of topics from visual and textual video data automatically and systematically. The proposed framework decodes the taxonomy of content efficiently through videos in different levels of noise and languages. empirical analysis of the YouTube-8M dataset shows the potential for the approach to change the paradigm in video advertising. Built to be scalable and easily adaptable, this solution can handle multifarious and complex user-generated content well, suited for a wide range of applications across various media platforms.
暂无评论