Thailand

The large size of pretrained networks makes them difficult to deploy for multiple tasks in storage-constrained settings.
Diff pruning enables parameter-efficient transfer learning that scales well with new tasks. The approach learns a task-specific &quot;diff&quot; vector that extends the original pretrained parameters. This diff vector is adaptively pruned during training with a differentiable approximation to the L0-norm penalty to encourage sparsity. As the number of tasks increases, diff pruning remains parameter-efficient, as it requires storing only a small diff vector for each task. Since it does not require access to all tasks during training, it is attractive in on-device deployment settings where tasks arrive in stream or even from different providers. Diff pruning can match the performance of finetuned baselines on the GLUE benchmark while only modifying 0.5% of the pretrained model&#39;s parameters per task and scales favorably in comparison to popular pruning approaches.

IJCNLP-AACL 2021

Parameter-Efficient Transfer Learning with Diff Pruning

efficient nlp; transfer learning; finetuning

**Welcome to ACL-IJCNLP 2021!**

The great event is jointly organized by the Association for Computational Linguistics (ACL) and Asian Federation of Natural Language Processing (AFNLP). 

As in previous years, the program of the conference includes a poster session, tutorials, workshops and demonstrations in addition to the main conference.


We were able to keep the registration fees similar to those charged for the virtual ACL 2020. The one fee allows attendance at the main conference and any/all tutorials and workshops. These fees would be $125 Regular Early and $175 Regular Late; $50 Student Early and $75 Student Late. Early registration closes at midnight July 11, 2021 (Eastern Daylight Time).

**Reminder:** It is ACL’s policy that at least one author of each accepted paper (including ACL Finding papers) must register for the conference.

**Reminder2:** Underline site will open closer to the event. If you already registered you will receive access detail

Registration is now open

The great event is jointly organized by the Association for Computational Linguistics (ACL) and Asian Federation of Natural Language Processing (AFNLP).

The large size of pretrained networks makes them difficult to deploy for multiple tasks in storage-constrained settings.
Diff pruning enables parameter-efficient transfer learning that scales well with new tasks. The approach learns a task-specific "diff" vector that extends the original pretrained parameters. This diff vector is adaptively pruned during training with a differentiable approximation to the L0-norm penalty to encourage sparsity. As the number of tasks increases, diff pruning remains parameter-efficient, as it requires storing only a small diff vector for each task. Since it does not require access to all tasks during training, it is attractive in on-device deployment settings where tasks arrive in stream or even from different providers. Diff pruning can match the performance of finetuned baselines on the GLUE benchmark while only modifying 0.5% of the pretrained model's parameters per task and scales favorably in comparison to popular pruning approaches.

technical paper

We study the problem of event coreference resolution (ECR) that seeks to group coreferent event mentions into the same clusters. Deep learning methods have recently been applied for this task to deliver state-of-the-art performance. However, existing deep learning models for ECR are limited in that they cannot exploit important interactions between relevant objects for ECR, e.g., context words and entity mentions, to support the encoding of document-level context. In addition, consistency constraints between golden and predicted clusters of event mentions have not been considered to improve representation learning in prior deep learning models for ECR. This work addresses such limitations by introducing a novel deep learning model for ECR. At the core of our model are document structures to explicitly capture relevant objects for ECR. Our document structures introduce diverse knowledge sources (discourse, syntax, semantics) to compute edges/interactions between structure nodes for document-level representation learning. We also present novel regularization techniques based on consistencies of golden and predicted clusters for event mentions in documents. Extensive experiments show that our model achieve state-of-the-art performance on two benchmark datasets.

Exploiting Document Structures and Cluster Consistencies for Event Coreference Resolution

Relational triple extraction is critical to understanding massive text corpora and constructing large-scale knowledge graph, which has attracted increasing research interest. However, existing studies still face some challenging issues, including information loss, error propagation and ignoring the interaction between entity and relation. To intuitively explore the above issues and address them, in this paper, we provide a revealing insight into relational triple extraction from a stereoscopic perspective, which rationalizes the occurrence of these issues and exposes the shortcomings of existing methods. Further, a novel model is proposed for relational triple extraction, which maps relational triples to a three-dimension (3-D) space and leverages three decoders to extract them, aimed at simultaneously handling the above issues. A series of experiments are conducted on five public datasets, demonstrating that the proposed model outperforms the recent advanced baselines.

StereoRel: Relational Triple Extraction from a Stereoscopic Perspective

Identifying causal relations of events is an important task in natural language processing area. However, the task is very challenging, because event causality is usually expressed in diverse forms that often lack explicit causal clues. Existing methods cannot handle well the problem, especially in the condition of lacking training data. Nonetheless, humans can make a correct judgement based on their background knowledge, including descriptive knowledge and relational knowledge. Inspired by it, we propose a novel Latent Structure Induction Network (LSIN) to incorporate the external structural knowledge into this task. Specifically, to make use of the descriptive knowledge, we devise a Descriptive Graph Induction module to obtain and encode the graph-structured descriptive knowledge. To leverage the relational knowledge, we propose a Relational Graph Induction module which is able to automatically learn a reasoning structure for event causality reasoning. Experimental results on two widely used datasets indicate that our approach significantly outperforms previous state-of-the-art methods.

Knowledge-Enriched Event Causality Identification via Latent Structure Induction Networks

Recent studies show that neural natural language processing (NLP) models are vulnerable to backdoor attacks. Injected with backdoors, models perform normally on benign examples but produce attacker-specified predictions when the backdoor is activated, presenting serious security threats to real-world applications. Since existing textual backdoor attacks pay little attention to the invisibility of backdoors, they can be easily detected and blocked. In this work, we present invisible backdoors that are activated by a learnable combination of word substitution. We show that NLP models can be injected with backdoors that lead to a nearly 100% attack success rate, whereas being highly invisible to existing defense strategies and even human inspections. The results raise a serious alarm to the security of NLP models, which requires further research to be resolved. All the data and code of this paper are released at https://github.com/thunlp/BkdAtk-LWS.

Turn the Combination Lock: Learnable Textual Backdoor Attacks via Word Substitution

Recent papers have introduced methods to incorporate gazetteer features and entity segmentation techniques in neural named entity recognition models. These papers rely on different resources and 
include features not related to the use of gazetteers, rendering impossible the comparison of the relative effectiveness of the approaches. Here, we provide a comprehensive overview of methods for incorporating gazetteers and for entity segmentation. We evaluate representative methods from each in similar settings for a fair comparison and identify the ones that are consistently better across datasets and input representations.
We further show that gazetteers improve entity segmentation and not just entity typing. Hence, we explore their utility in recognizing long entities, a problem for which entity segmentation techniques were developed.
Our work explains the mechanisms via which gazetteers improve the performance of neural NER models.

The Utility and Interplay of Gazetteers and Entity Segmentation for Named Entity Recognition in English

Neural network algorithms such as those based on transformers and attention models have excelled on Automatic Text Classification (ATC) tasks. However, such enhanced performance comes at high computational costs. Ensembles of simpler classifiers (i.e., stacking) that exploit algorithmic and representational complementarities have also been shown to produce top-notch performance in ATC, enjoying high effectiveness and potentially lower computational costs. 

In this context, we present the first and largest comparative study to exploit the cost-effectiveness of stacking of ATC classifiers, consisting of transformers and non-neural algorithms. In particular, we are interested in answering research questions such as: Is it possible to obtain an effective ensemble with significantly less computational cost than the best learning model for a given dataset? Disregarding the computational cost, can an ensemble improve the effectiveness of the best learning model? Besides answering such questions, another main contribution of this paper is the proposal of a low-cost oracle-based method that can predict the best ensemble in each scenario (with and without computational cost limitations) using only a fraction of the available training data.

On the Cost-Effectiveness of Stacking of Neural and Non-Neural Methods for Text Classification: Scenarios and Performance Prediction

In intensive care units (ICUs), patient health is monitored through (1) continuous vital signals from various medical devices, and (2) clinical notes consisting of opinions and summaries from doctors which are recorded in electronic health records (EHR). It is difficult to jointly model these two sources of information because clinical notes, unlike vital signals, are collected at irregular intervals and their contents are relatively unstructured. In this paper, we present a model that combines both sources of information about ICU patients to make accurate in-hospital mortality predictions. 
We apply a fine-tuned BERT model to each of the patient's clinical notes. The resulting embeddings are then combined to obtain the overall embedding for the entire text part of the data. This is then combined with the output of an LSTM model that encodes patients' vital signals.
Our model improves upon the state of the art for mortality prediction, attaining an AUC score of 0.9, compared to the previous 0.87, setting a new standard for mortality prediction on the MIMIC III benchmark.

Predicting in-hospital mortality by combining clinical notes with time-series data

We computationally model the processes of word borrowing from a donor word to an incorporated word, and vice versa, by answering two questions: (1) what does a word look like if it is incorporated into another language, and in the opposite direction (2) where did a word come from? We employ neural sequence models, focusing on six specific borrowing relations: calques, partial calques, semantic loans, phono-semantic matches, transliterations, and generic borrowings. We experiment with several model variants, including LSTM encoder-decoders, copy attention, and Transformers, along with several data distribution and preprocessing scenarios. We find that the quality of data strongly influences the performance of models, with low character edit distance between model predictions and the gold words.

Sequence Models for Computational Etymology of Borrowings

Cross-lingual word embeddings provide a way
for information to be transferred between languages. In this paper we evaluate an extension of a joint training approach to learning cross-lingual embeddings that incorporates
sub-word information during training. This
method could be particularly well-suited to
lower-resource and morphologically-rich languages because it can be trained on modest
size monolingual corpora, and is able to represent out-of-vocabulary words (OOVs). We
consider bilingual lexicon induction, including
an evaluation focused on OOVs. We find that
this method achieves improvements over previous approaches, particularly for OOVs.

Evaluating a Joint Training Approach for Learning Cross-lingual Embeddings with Sub-word Information without Parallel Corpora on Lower-resource Languages

Abductive reasoning starts from some observations and aims at finding the most plausible explanation for these observations. To perform abduction, humans often make use of temporal and causal inferences and knowledge about how some hypothetical situation can result in different outcomes. This work offers the first study of how such knowledge impacts the Abductive NLI task -- which consists in choosing the more likely explanation for given observations. We train a specialized language model LMI that is tasked to generate what could happen next from a hypothetical scenario that evolves from a given event. We then propose a multi-task model MTL to solve the Abductive NLI task, which predicts a plausible explanation by a) considering different possible events emerging from candidate hypotheses -- events generated by LMI -- and b) selecting the one that is most similar to the observed outcome. We show that our MTL model improves over prior vanilla pre-trained LMs fine-tuned on Abductive NLI. Our manual evaluation and analysis suggest that learning about possible next events from different hypothetical scenarios supports abductive inference.

Parameter-Efficient Transfer Learning with Diff Pruning

Downloads

Next from IJCNLP-AACL 2021

Exploiting Document Structures and Cluster Consistencies for Event Coreference Resolution

Similar lecture

PhotoChat: A Human-Human Dialogue Dataset With Photo Sharing Behavior For Joint Image-Text Modeling

Parameter-Efficient Transfer Learning with Diff Pruning

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from IJCNLP-AACL 2021

Exploiting Document Structures and Cluster Consistencies for Event Coreference Resolution

Similar lecture

PhotoChat: A Human-Human Dialogue Dataset With Photo Sharing Behavior For Joint Image-Text Modeling

Downloads