Search Results for author:
Found 40 papers, 19 papers with code
Model Editing by Pure Fine-Tuning
Fine-tuning is dismissed as not effective for model editing due to its poor performance compared to more specialized methods.
Seq2seq is All You Need for Coreference Resolution
Existing works on coreference resolution suggest that task-specific models are necessary to achieve state-of-the-art performance.
Improving Multitask Retrieval by Promoting Task Specialization
In multitask retrieval, a single retriever is trained to retrieve relevant contexts for multiple tasks.
EntQA: Entity Linking as Question Answering
A conventional approach to entity linking is to first find mentions in a given document and then infer their underlying entities in the knowledge base.
Ranked #4 on
Entity Linking
on AIDA-CoNLL
Understanding Hard Negatives in Noise Contrastive Estimation
The choice of negative examples is important in noise contrastive estimation.
Fast and Effective Biomedical Entity Linking Using a Dual Encoder
Additionally, we modify our dual encoder model for end-to-end biomedical entity linking that performs both mention span detection and entity disambiguation and out-performs two recently proposed models.
Data-to-text Generation by Splicing Together Nearest Neighbors
We propose to tackle data-to-text generation tasks by directly splicing together retrieved segments of text from "neighbor" source-target pairs.
Corrected CBOW Performs as well as Skip-gram
Mikolov et al. (2013a) observed that continuous bag-of-words (CBOW) word embeddings tend to underperform Skip-gram (SG) embeddings, and this finding has been reported in subsequent works.
Unsupervised Label Refinement Improves Dataless Text Classification
This reliance causes dataless classifiers to be highly sensitive to the choice of label descriptions and hinders the broader application of dataless classification in practice.
Ranked #3 on
Zero-Shot Text Classification
on AG News
Mining Knowledge for Natural Language Inference from Wikipedia Categories
Accurate lexical entailment (LE) and natural language inference (NLI) often require large quantities of costly annotations.
NatCat: Weakly Supervised Text Classification with Naturally Annotated Resources
We describe NatCat, a large-scale resource for text classification constructed from three data sources: Wikipedia, Stack Exchange, and Reddit.
Discrete Latent Variable Representations for Low-Resource Text Classification
While much work on deep latent variable models of text uses continuous latent variables, discrete latent variables are interesting because they are more interpretable and typically more space efficient.
Learning Discrete Structured Representations by Adversarially Maximizing Mutual Information
We propose learning discrete structured representations from unlabeled data by maximizing the mutual information between a structured latent variable and a target variable.
EntEval: A Holistic Evaluation Benchmark for Entity Representations
Rich entity representations are useful for a wide class of problems involving entities.
Label-Agnostic Sequence Labeling by Copying Nearest Neighbors
Retrieve-and-edit based approaches to structured prediction, where structures associated with retrieved neighbors are edited to form new structures, have recently attracted increased interest.
Formal Limitations on the Measurement of Mutual Information
Measuring mutual information from finite data is difficult.
Compositional Morpheme Embeddings with Affixes as Functions and Stems as Arguments
StAffNet, the name of our architecture, shows competitive performance with the state-of-the-art on this task.
Mutual Information Maximization for Simple and Accurate Part-Of-Speech Induction
We address part-of-speech (POS) induction by maximizing the mutual information between the induced label and its context.
OneNet: Joint Domain, Intent, Slot Prediction for Spoken Language Understanding
In practice, most spoken language understanding systems process user input in a pipelined manner; first domain is predicted, then intent and semantic slots are inferred according to the semantic frames of the predicted domain.
Reconstruction of Word Embeddings from Sub-Word Parameters
Pre-trained word embeddings improve the performance of a neural model at the cost of increasing the model size.
A Sub-Character Architecture for Korean Language Processing
We introduce a novel sub-character architecture that exploits a unique compositional structure of the Korean language.
Adversarial Adaptation of Synthetic or Stale Data
Both cause a distribution mismatch between training and evaluation, leading to a model that overfits the flawed training data and performs poorly on the test data.
Domain Attention with an Ensemble of Experts
When given domain K + 1, our model uses a weighted combination of the K domain experts{'} feedback along with its own opinion to make predictions on the new domain.
Entity Identification as Multitasking
Standard approaches in entity identification hard-code boundary detection and type prediction into labels (e. g., John/B-PER Smith/I-PER) and then perform Viterbi.
Domainless Adaptation by Constrained Decoding on a Schema Lattice
In many applications such as personal digital assistants, there is a constant need for new domains to increase the system{'}s coverage of user queries.
Frustratingly Easy Neural Domain Adaptation
Popular techniques for domain adaptation such as the feature augmentation method of Daum{\'e} III (2009) have mostly been considered for sparse binary-valued features, but not for dense real-valued features such as those used in neural networks.
Unsupervised Part-Of-Speech Tagging with Anchor Hidden Markov Models
We tackle unsupervised part-of-speech (POS) tagging by learning hidden Markov models (HMMs) that are particularly well-suited for the problem.
