Papers Data Ablation
“Data Ablation” 태그가 달린 논문 10편 · 필터 해제
Toward Stable and Consistent Evaluation Results: A New Methodology for Base Model Evaluation
This paper poses two critical issues in evaluating base models (without post-training): (1) Unstable evaluation during training: in the early stages of pre-training, the models lack the capability to answer questions as …
Data AblationScalable Data Ablation Approximations for Language Models through Modular Training and Merging
Training data compositions for Large Language Models (LLMs) can significantly affect their downstream performance. However, a thorough data ablation study exploring large sets of candidate data mixtures is typically proh…
Data AblationUnleashing the Power of LLMs as Multi-Modal Encoders for Text and Graph-Structured Data
Graph-structured information offers rich contextual information that can enhance language models by providing structured relationships and hierarchies, leading to more expressive embeddings for various applications such …
Contrastive LearningData AblationQuestion AnsweringRetrieval+1Alzheimer's disease detection in PSG signals
Alzheimer's disease (AD) and sleep disorders exhibit a close association, where disruptions in sleep patterns often precede the onset of Mild Cognitive Impairment (MCI) and early-stage AD. This study delves into the pote…
Alzheimer's Disease DetectionData AblationEEGMulti-horizon short-term load forecasting using hybrid of LSTM and modified split convolution
Precise short-term load forecasting (STLF) plays a crucial role in the smooth operation of power systems, future capacity planning, unit commitment, and demand response. However, due to its non-stationary and its dep…
Data AblationLoad ForecastingMissing ElementsMultivariate Time Series Forecasting+3Understanding Robust Overfitting of Adversarial Training and Beyond
Robust overfitting widely exists in adversarial training of deep networks. The exact underlying reasons for this are still not completely understood. Here, we explore the causes of robust overfitting by comparing the dat…
Adversarial RobustnessData AblationLocalize, Group, and Select: Boosting Text-VQA by Scene Text Modeling
As an important task in multimodal context understanding, Text-VQA (Visual Question Answering) aims at question answering through reading text information in images. It differentiates from the original VQA task as Text-V…
Data AblationOptical Character RecognitionOptical Character Recognition (OCR)Question Answering+3A comprehensive empirical analysis on cross-domain semantic enrichment for detection of depressive language
We analyze the process of creating word embedding feature representations designed for a learning task when annotated data is scarce, for example, in depressive language detection from Tweets. We start with a rich word e…
Data AblationPersonalized Embedding-based e-Commerce Recommendations at eBay
Recommender systems are an essential component of e-commerce marketplaces, helping consumers navigate massive amounts of inventory and find what they need or love. In this paper, we present an approach for generating per…
Data AblationNavigateRecommendation SystemsAI Playground: Unreal Engine-based Data Ablation Tool for Deep Learning
Machine learning requires data, but acquiring and labeling real-world data is challenging, expensive, and time-consuming. More importantly, it is nearly impossible to alter real data post-acquisition (e.g., change the il…
Data AblationDepth EstimationIndoor Monocular Depth EstimationSemantic Segmentation+1