An official website of the United States government
Does prompting a large language model (LLM) like GPT-3 with explanations improve in-context learning? We study this question on two NLP tasks that involve reasoning over text, namely question answering and natural language inference. We test the performance of four LLMs on three textual reasoning datasets using prompts that include explanations in multiple different styles. For these tasks, we find that including explanations in the prompts for OPT, GPT-3 (davinci), and InstructGPT (text-davinci-001) only yields small to moderate accuracy improvements over standard few-show learning. However, text-davinci-002 is able to benefit more substantially. We further show that explanations generated by the LLMs may not entail the models' predictions nor be factually grounded in the input, even on simple tasks with extractive explanations. However, these flawed explanations can still be useful as a way to verify LLMs' predictions post-hoc. Through analysis in our three settings, we show that explanations judged by humans to be good---logically consistent with the input and the prediction---more likely cooccur with accurate predictions. Following these observations, we train calibrators using automatically extracted scores that assess the reliability of explanations, allowing us to improve performance post-hoc across all of our datasets.
more »
« less
Jointly Identifying and Fixing Inconsistent Readings from Information Extraction Systems
Information extraction systems analyze text to produce entities and beliefs, but their output often has errors. In this paper we analyze the reading consistency of the extracted facts with respect to the text from which they were derived and show how to detect and correct errors. We consider both the scenario when the provenance text is automatically found by an IE system and when it is curated by humans. We contrast consistency with credibility; define and explore consistency and repair tasks; and demonstrate a simple, yet effective and generalizable, model. We analyze these tasks and evaluate this approach on three datasets. Against a strong baseline model, we consistently improve both consistency and repair across three datasets using a simple MLP model with attention and lexical features.
more »
« less
- PAR ID:
- 10382769
- Date Published:
- Journal Name:
- Third Deep Learning Inside Out (DeeLIO) Workshop: Knowledge Extraction and Integration for Deep Learning Architecture
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Recent advances in natural language processing (NLP) and Big Data technologies have been crucial for scientists to analyze political unrest and violence, prevent harm, and promote global conflict management. Government agencies and public security organizations have invested heavily in deep learning-based applications to study global conflicts and political violence. However, such applications involving text classification, information extraction, and other NLP-related tasks require extensive human efforts in annotating/labeling texts. While limited labeled data may drastically hurt the models’ performance (over-fitting), large demands on annotation tasks may turn real-world applications impracticable. To address this problem, we propose Confli-T5, a prompt-based method that leverages the domain knowledge from existing political science ontology to generate synthetic but realistic labeled text samples in the conflict and mediation domain. Our model allows generating textual data from the ground up and employs our novel Double Random Sampling mechanism to improve the quality (coherency and consistency) of the generated samples. We conduct experiments over six standard datasets relevant to political science studies to show the superiority of Confli-T5. Our codes are publicly availablemore » « less
-
While neural networks are used for classification tasks across domains, a long-standing open problem in machine learning is determining whether neural networks trained using standard procedures are consistent for classification, i.e., whether such models minimize the probability of misclassification for arbitrary data distributions. In this work, we identify and construct an explicit set of neural network classifiers that are consistent. Since effective neural networks in practice are typically both wide and deep, we analyze infinitely wide networks that are also infinitely deep. In particular, using the recent connection between infinitely wide neural networks and neural tangent kernels, we provide explicit activation functions that can be used to construct networks that achieve consistency. Interestingly, these activation functions are simple and easy to implement, yet differ from commonly used activations such as ReLU or sigmoid. More generally, we create a taxonomy of infinitely wide and deep networks and show that these models implement one of three well-known classifiers depending on the activation function used: 1) 1-nearest neighbor (model predictions are given by the label of the nearest training example); 2) majority vote (model predictions are given by the label of the class with the greatest representation in the training set); or 3) singular kernel classifiers (a set of classifiers containing those that achieve consistency). Our results highlight the benefit of using deep networks for classification tasks, in contrast to regression tasks, where excessive depth is harmful.more » « less
-
Abstract Text augmentation is an effective technique in alleviating overfitting in NLP tasks. In existing methods, text augmentation and downstream tasks are mostly performed separately. As a result, the augmented texts may not be optimal to train the downstream model. To address this problem, we propose a three-level optimization framework to perform text augmentation and the downstream task end-to- end. The augmentation model is trained in a way tailored to the downstream task. Our framework consists of three learning stages. A text summarization model is trained to perform data augmentation at the first stage. Each summarization example is associated with a weight to account for its domain difference with the text classification data. At the second stage, we use the model trained at the first stage to perform text augmentation and train a text classification model on the augmented texts. At the third stage, we evaluate the text classification model trained at the second stage and update weights of summarization examples by minimizing the validation loss. These three stages are performed end-to-end. We evaluate our method on several text classification datasets where the results demonstrate the effectiveness of our method. Code is available at https://github.com/Sai-Ashish/End-to-End-Text-Augmentation.more » « less
-
Vedaldi, Andrea; Bischof, Horst; Brox, Thomas; Frahm, Jan-Michael (Ed.)Novel view video synthesis aims to synthesize novel viewpoints videos given input captures of a human performance taken from multiple reference viewpoints and over consecutive time steps. Despite great advances in model-free novel view synthesis, existing methods present three limitations when applied to complex and time-varying human performance. First, these methods (and related datasets) mainly consider simple and symmetric objects. Second, they do not enforce explicit consistency across generated views. Third, they focus on static and non-moving objects. The fine-grained details of a human subject can therefore suffer from inconsistencies when synthesized across different viewpoints or time steps. To tackle these challenges, we introduce a human-specific framework that employs a learned 3D-aware representation. Specifically, we first introduce a novel siamese network that employs a gating layer for better reconstruction of the latent volumetric representation and, consequently, final visual results. Moreover, features from consecutive time steps are shared inside the network to improve temporal consistency. Second, we introduce a novel loss to explicitly enforce consistency across generated views both in space and in time. Third, we present the Multi-View Human Action (MVHA) dataset, consisting of near 1200 synthetic human performance captured from 54 viewpoints. Experiments on the MVHA, Pose-Varying Human Model and ShapeNet datasets show that our method outperforms the state-of-the-art baselines both in view generation quality and spatio-temporal consistency.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.18653/v1/2022.deelio-1.5
