Recent research in mechanistic interpretability has attempted to
reverse...
Fine-tuning language models (LMs) has yielded success on diverse downstr...
Semantic textual similarity (STS) has been a cornerstone task in NLP tha...
Retrieval-based language models (LMs) have demonstrated improved
interpr...
Transformer-based language models (LMs) are powerful and widely-applicab...
Large language models (LLMs) have emerged as a widely-used tool for
info...
Large language models (LLMs) exploit in-context learning (ICL) to solve ...
We consider the task of text generation in language models with constrai...
Scaling up language models has led to unprecedented performance gains, b...
Pre-trained language models encode undesirable social biases, which are
...
Masked language models like BERT can perform text classification in a
ze...
Many NLP datasets have been found to contain shortcuts: simple decision ...
It has become standard to solve NLP tasks by fine-tuning pre-trained lan...
Pre-trained masked language models successfully perform few-shot learnin...
Dense retrieval uses a contrastive learning framework to learn dense
rep...
Recent work has improved language models remarkably by equipping them wi...
Deductive reasoning (drawing conclusions from assumptions) is a challeng...
Federated learning allows distributed users to collaboratively train a m...
A growing line of work has investigated the development of neural NLP mo...
The growing size of neural language models has led to increased attentio...
Masked language models conventionally use a masking rate of 15
belief th...
Conversational question answering (CQA) systems aim to provide
natural-l...
Many datasets have been created for training reading comprehension model...
Open-domain question answering has exploded in popularity recently due t...
Dense retrieval methods have shown great promise over sparse retrieval
m...
Word sense disambiguation (WSD) is a long-standing problem in natural
la...
This paper presents SimCSE, a simple contrastive learning framework that...
Petroni et al. (2019) demonstrated that it is possible to retrieve world...
We review the EfficientQA competition from NeurIPS 2020. The competition...
The recent GPT-3 model (Brown et al., 2020) achieves remarkable few-shot...
Open-domain question answering can be reformulated as a phrase retrieval...
End-to-end relation extraction aims to identify named entities and extra...
An unsolved challenge in distributed or federated learning is to effecti...
Open-domain question answering relies on efficient passage retrieval to
...
This paper presents a general approach for open-domain question answerin...
We present the results of the Machine Reading for Question Answering (MR...
Many question answering (QA) tasks only provide weak supervision for how...
Language model pretraining has led to significant performance gains but
...
We present SpanBERT, a pre-training method that is designed to better
re...
Humans gather information by engaging in conversations involving a serie...
This paper proposes to tackle open- domain question answering using Wiki...
Enabling a computer to understand a document so that it can answer
compr...
Knowledge bases provide applications with the benefit of easily accessib...