Search Results for author:
Found 22 papers, 15 papers with code
RA-DIT: Retrieval-Augmented Dual Instruction Tuning
Retrieval-augmented language models (RALMs) improve performance by accessing long-tail and up-to-date knowledge from external data stores, but are challenging to build.
xSIM++: An Improved Proxy to Bitext Mining Performance for Low-Resource Languages
In comparison to xSIM, we show that xSIM++ is better correlated with the downstream BLEU scores of translation systems trained on mined bitexts, providing a reliable proxy of bitext mining performance without needing to run expensive bitext mining pipelines.
Few-Shot Data Synthesis for Open Domain Multi-Hop Question Answering
Few-shot learning for open domain multi-hop question answering typically relies on the incontext learning capability of large language models (LLMs).
BLASER: A Text-Free Speech-to-Speech Translation Evaluation Metric
In this paper, we propose a text-free evaluation metric for end-to-end S2ST, named BLASER, to avoid the dependency on ASR systems.
Automatic Speech Recognition
Automatic Speech Recognition (ASR)
+3
Leveraging Natural Supervision for Language Representation Learning and Generation
In this thesis, we describe three lines of work that seek to improve the training and evaluation of neural models using naturally-occurring supervision.
Improving In-Context Few-Shot Learning via Self-Supervised Training
Self-supervised pretraining has made few-shot learning possible for many NLP tasks.
TVStoryGen: A Dataset for Generating Stories with Character Descriptions
We introduce TVStoryGen, a story generation dataset that requires generating detailed TV show episode recaps from a brief summary and a set of documents describing the characters involved.
Ranked #1 on
Story Generation
on Fandom dev
SummScreen: A Dataset for Abstractive Screenplay Summarization
Since characters are fundamental to TV series, we also propose two entity-centric evaluation metrics.
WikiTableT: A Large-Scale Data-to-Text Dataset for Generating Wikipedia Article Sections
Datasets for data-to-text generation typically focus either on multi-domain, single-sentence generation or on single-domain, long-form generation.
Exemplar-Controllable Paraphrasing and Translation using Bitext
Our experimental results show that our models achieve competitive results on controlled paraphrase generation and strong performance on controlled machine translation.
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.
A novel random access scheme for M2M communication in crowded asynchronous massive MIMO systems
Simulation results show that, compared to the exiting random access scheme for the crowded asynchronous massive MIMO systems, the proposed scheme can improve the uplink throughput and estimate the effective timing offsets accurately at the same time.
Learning Probabilistic Sentence Representations from Paraphrases
Probabilistic word embeddings have shown effectiveness in capturing notions of generality and entailment, but there is very little work on doing the analogous type of investigation for sentences.
How to Ask Better Questions? A Large-Scale Multi-Domain Dataset for Rewriting Ill-Formed Questions
We present a large-scale dataset for the task of rewriting an ill-formed natural language question to a well-formed one.
ALBERT: A Lite BERT for Self-supervised Learning of Language Representations
Increasing model size when pretraining natural language representations often results in improved performance on downstream tasks.
Ranked #1 on
Natural Language Inference
on QNLI
EntEval: A Holistic Evaluation Benchmark for Entity Representations
Rich entity representations are useful for a wide class of problems involving entities.
Evaluation Benchmarks and Learning Criteria for Discourse-Aware Sentence Representations
Prior work on pretrained sentence embeddings and benchmarks focus on the capabilities of stand-alone sentences.
Variational Sequential Labelers for Semi-Supervised Learning
Our model family consists of a latent-variable generative model and a discriminative labeler.
Ranked #72 on
Named Entity Recognition (NER)
on CoNLL 2003 (English)
Smaller Text Classifiers with Discriminative Cluster Embeddings
Word embedding parameters often dominate overall model sizes in neural methods for natural language processing.
Controllable Paraphrase Generation with a Syntactic Exemplar
Prior work on controllable text generation usually assumes that the controlled attribute can take on one of a small set of values known a priori.
Variational recurrent models for representation learning
Existing variational recurrent models typically use stochastic recurrent connections to model the dependence among neighboring latent variables, while generation assumes independence of generated data per time step given the latent sequence.
A Multi-Task Approach for Disentangling Syntax and Semantics in Sentence Representations
We propose a generative model for a sentence that uses two latent variables, with one intended to represent the syntax of the sentence and the other to represent its semantics.
