Search Results for author:
Found 21 papers, 8 papers with code
SELF-[IN]CORRECT: LLMs Struggle with Refining Self-Generated Responses
Can LLMs continually improve their previous outputs for better results?
FollowIR: Evaluating and Teaching Information Retrieval Models to Follow Instructions
We introduce our dataset FollowIR, which contains a rigorous instruction evaluation benchmark as well as a training set for helping IR models learn to better follow real-world instructions.
Dated Data: Tracing Knowledge Cutoffs in Large Language Models
Using this analysis, we find that effective cutoffs often differ from reported cutoffs.
Enhancing Systematic Decompositional Natural Language Inference Using Informal Logic
Contemporary language models enable new opportunities for structured reasoning with text, such as the construction and evaluation of intuitive, proof-like textual entailment trees without relying on brittle formal logic.
When do Generative Query and Document Expansions Fail? A Comprehensive Study Across Methods, Retrievers, and Datasets
Using large language models (LMs) for query or document expansion can improve generalization in information retrieval.
MegaWika: Millions of reports and their sources across 50 diverse languages
To foster the development of new models for collaborative AI-assisted report generation, we introduce MegaWika, consisting of 13 million Wikipedia articles in 50 diverse languages, along with their 71 million referenced source materials.
"According to ...": Prompting Language Models Improves Quoting from Pre-Training Data
Large Language Models (LLMs) may hallucinate and generate fake information, despite pre-training on factual data.
NevIR: Negation in Neural Information Retrieval
Although the Information Retrieval (IR) community has adopted LMs as the backbone of modern IR architectures, there has been little to no research in understanding how negation impacts neural IR.
Synthetic Cross-language Information Retrieval Training Data
By repeating this process, collections of arbitrary size can be created in the style of MS MARCO but using naturally-occurring documents in any desired genre and domain of discourse.
When Do Decompositions Help for Machine Reading?
Answering complex questions often requires multi-step reasoning in order to obtain the final answer.
Defending Against Disinformation Attacks in Open-Domain Question Answering
Recent work in open-domain question answering (ODQA) has shown that adversarial poisoning of the search collection can cause large drops in accuracy for production systems.
Pretrained Models for Multilingual Federated Learning
Since the advent of Federated Learning (FL), research has applied these methods to natural language processing (NLP) tasks.
When to Use Multi-Task Learning vs Intermediate Fine-Tuning for Pre-Trained Encoder Transfer Learning
We find that there is a simple heuristic for when to use one of these techniques over the other: pairwise MTL is better than STILTs when the target task has fewer instances than the supporting task and vice versa.
End-to-End Speech Translation for Code Switched Speech
Code switching (CS) refers to the phenomenon of interchangeably using words and phrases from different languages.
Exploring the Relationship Between Algorithm Performance, Vocabulary, and Run-Time in Text Classification
Text classification is a significant branch of natural language processing, and has many applications including document classification and sentiment analysis.
Streaming Models for Joint Speech Recognition and Translation
However, all previous work has only looked at this problem from the consecutive perspective, leaving uncertainty on whether these approaches are effective in the more challenging streaming setting.
Automatic Speech Recognition
Automatic Speech Recognition (ASR)
+2
Learning from Task Descriptions
Typically, machine learning systems solve new tasks by training on thousands of examples.
Can Humor Prediction Datasets be used for Humor Generation? Humorous Headline Generation via Style Transfer
Understanding and identifying humor has been increasingly popular, as seen by the number of datasets created to study humor.
You Don't Have Time to Read This: An Exploration of Document Reading Time Prediction
Predicting reading time has been a subject of much previous work, focusing on how different words affect human processing, measured by reading time.
The rJokes Dataset: a Large Scale Humor Collection
We also introduce this dataset as a task for future work, where models learn to predict the level of humor in a joke.
Humor Detection: A Transformer Gets the Last Laugh
These experiments show that this method outperforms all previous work done on these tasks, with an F-measure of 93. 1% for the Puns dataset and 98. 6% on the Short Jokes dataset.
