Dominican Republic

Recent pretrained language models extend from millions to billions of parameters. Thus the need to fine-tune an extremely large pretrained model with a limited training corpus arises in various downstream tasks. In this paper, we propose a straightforward yet effective fine-tuning technique, ChildTuning, which updates a subset of parameters (called child network) of large pretrained models via strategically masking out the gradients of the non-child network during the backward process. Experiments on various downstream tasks in GLUE benchmark show that ChildTuning consistently outperforms the vanilla fine-tuning by 1.5~8.6 average score among four different pretrained models, and surpasses the prior fine-tuning techniques by 0.6~1.3 points. Furthermore, empirical results on domain transfer and task transfer show that ChildTuning can obtain better generalization performance by large margins.

EMNLP 2021

Raise a Child in Large Language Model: Towards Effective and Generalizable Fine-tuning

EMNLP 2021 is planned to be a hybrid event in Punta Cana, Dominican Republic, with both on-site and fully virtual participation possible. The experience for on-site participants would closely approximate a normal pre-COVID *ACL conference, with 5-6 thematically organized parallel sessions and live Q/A and interactive discussion immediately after the talks. Presentations by virtual participants will be equitably interleaved with those of on-site participants, projected on the auditorium screens as if on-site, and also followed immediately by live Q/A and interactive discussion at a time during reasonable waking hours for the virtual presenter. For all participants, on-site and virtual, who are unable to attend a session due to either time-zone issues or because they are participating in another session live, talk recordings and slides will be available online at a minimum after the live presentation (and in many cases before as well), and questions may be submitted in advance on session-specific discussion boards and answered live in session with the usual visual aids if desired.

<iframe style="width:700px;height:400px" src="https://online.fliphtml5.com/ebtyf/ceby/"  seamless="seamless" scrolling="no" frameborder="0" allowtransparency="true" allowfullscreen="true" ></iframe>

Please Note: The EMNLP registration system is not currently connected to the underline site as we are still in the process of building out EMNLP 2021. You will receive access instructions from underline the week of November 1st. 

Access is given only to EMNLP upon registration, if you have not registered please do so [here](https://2021.emnlp.org/registration).

Registered attendees will receive access the week of November 1st.

EMNLP 2021 is planned to be a hybrid event in Punta Cana, Dominican Republic, with both on-site and fully virtual participation possible.

poster

Transformer models are permutation equivariant. To supply the order and type information of the input tokens, position and segment embeddings are usually added to the input. Recent works proposed variations of positional encodings with relative position encodings achieving better performance. Our analysis shows that the gain actually comes from moving positional information to attention layer from the input. Motivated by this, we introduce Decoupled posItional attEntion for Transformers (DIET), a simple yet effective mechanism to encode position and segment information into the Transformer models. The proposed method has faster training and inference time, while achieving competitive performance on GLUE, XTREME and WMT benchmarks. We further generalize our method to long-range transformers and show performance gain.

A Simple and Effective Positional Encoding for Transformers

Are pairs of words that tend to occur together also likely to stand in a linguistic dependency? This empirical question is motivated by a long history of literature in cognitive science, psycholinguistics, and NLP. In this work we contribute an extensive analysis of the relationship between linguistic dependencies and statistical dependence between words. Improving on previous work, we introduce the use of large pretrained language models to compute contextualized estimates of the pointwise mutual information between words (CPMI). For multiple models and languages, we extract dependency trees which maximize CPMI, and compare to gold standard linguistic dependencies. Overall, we find that CPMI dependencies achieve an unlabelled undirected attachment score of at most $\approx 0.5$. While far above chance, and consistently above a non-contextualized PMI baseline, this score is generally comparable to a simple baseline formed by connecting adjacent words. We analyze which kinds of linguistic dependencies are best captured in CPMI dependencies, and also find marked differences between the estimates of the large pretrained language models, illustrating how their different training schemes affect the type of dependencies they capture.

Linguistic Dependencies and Statistical Dependence

Natural language relies on a finite lexicon to express an unbounded set of emerging ideas. One result of this tension is the formation of new compositions, such that existing linguistic units can be combined with emerging items into novel expressions. We develop a framework that exploits the cognitive mechanisms of chaining and multimodal knowledge to predict emergent compositional expressions through time. We present the syntactic frame extension model (SFEM) that draws on the theory of chaining and knowledge from “percept”, “concept”, and “language” to infer how verbs extend their frames to form new compositions with existing and novel nouns. We evaluate SFEM rigorously on the 1) modalities of knowledge and 2) categorization models of chaining, in a syntactically parsed English corpus over the past 150 years. We show that multimodal SFEM predicts newly emerged verb syntax and arguments substantially better than competing models using purely linguistic or unimodal knowledge. We find support for an exemplar view of chaining as opposed to a prototype view and reveal how the joint approach of multimodal chaining may be fundamental to the creation of literal and figurative language uses including metaphor and metonymy.

Predicting emergent linguistic compositions through time: Syntactic frame extension via multimodal chaining

Taxonomies of writing systems since Gelb (1952) have classified systems based on what the written symbols represent: if they represent words or morphemes, they are logographic; if syllables, syllabic; if segments, alphabetic; etc. Sproat (2000) and Rogers (2005) broke with tradition by splitting the logographic and phonographic aspects into two dimensions, with logography being graded rather than a categorical distinction. A system could be syllabic, and highly logographic; or alphabetic, and mostly non-logographic. This accords better with how writing systems actually work, but neither author proposed a method for measuring logography.
In this paper we propose a novel measure of the degree of logography that uses an attention- based sequence-to-sequence model trained to predict the spelling of a token from its pronun- ciation in context. In an ideal phonographic system, the model should need to attend to only the current token in order to compute how to spell it, and this would show in the attention matrix activations. In contrast, with a logographic system, where a given pronunciation might correspond to several different spellings, the model would need to attend to a broader context. The ratio of the activation outside the token and the total activation forms the basis of our measure. We compare this with a simple lexical measure, and an entropic measure, as well as several other neural models, and argue that on balance our attention-based measure accords best with intuition about how logographic various systems are.
Our work provides the first quantifiable measure of the notion of logography that accords with linguistic intuition and, we argue, provides better insight into what this notion means.

The Taxonomy of Writing Systems: How to Measure how Logographic a System is

In this presentation, I am going to present our Findings Paper on WIKIBIAS. Biases continue to be prevalent in modern text
and media, especially subjective bias – a special type of bias that introduces improper attitudes or presents a statement with the presupposition of truth. To tackle the problem of detecting and further mitigating subjective bias,
we introduce a manually annotated parallel
corpus WIKIBIAS with more than 4,000 sentence pairs from Wikipedia edits. This corpus
contains annotations towards both sentence-level bias types and token-level biased segments. We present systematic analyses of our
dataset and results achieved by a set of state-of-the-art baselines in terms of three tasks: bias
classification, tagging biased segments, and
neutralizing biased text. We find that current
models still struggle with detecting multi-span
biases despite their reasonable performances,
suggesting that our dataset can serve as a useful research benchmark. We also demonstrate
that models trained on our dataset can generalize well to multiple domains such as news and
political speeches.

WIKIBIAS: Detecting Multi-Span Subjective Biases in Language

While large-scale pretrained language models have been shown to learn effective linguistic representations for many NLP tasks, there remain many real-world contextual aspects of language that current approaches do not capture. For instance, consider a cloze-test “I enjoyed the _______ game this weekend”: the correct answer depends heavily on where the speaker is from, when the utterance occurred, and the speaker’s broader social milieu and preferences. Although language depends heavily on the geographical, temporal, and other social contexts of the speaker, these elements have not been incorporated into modern transformer-based language models. We propose a simple but effective approach to incorporate speaker social context into the learned representations of large-scale language models. Our method first learns dense representations of social contexts using graph representation learning algorithms and then primes language model pretraining with these social context representations. We evaluate our approach on geographically sensitive language modeling tasks and show a substantial improvement (more than 100% relative lift on MRR) compared to baselines

LMSOC: An Approach for Socially Sensitive Pretraining

Compliments and concerns in reviews are valuable for understanding users' shopping interests and their opinions with respect to specific aspects of certain items. Existing review-based recommenders favor large and complex language encoders that can only learn latent and uninterpretable text representations. They lack explicit user attention and item property modeling, which however could provide valuable information beyond the ability to recommend items. Therefore, we propose a tightly coupled two-stage approach, including an Aspect-Sentiment Pair Extractor (ASPE) and an Attention-Property-aware Rating Estimator (APRE). Unsupervised ASPE mines Aspect-Sentiment pairs (AS-pairs) and APRE predicts ratings using AS-pairs as concrete aspect-level evidence. Extensive experiments on seven real-world Amazon Review Datasets demonstrate that ASPE can effectively extract AS-pairs which enable APRE to deliver superior accuracy over the leading baselines.

Recommend for a Reason: Unlocking the Power of Unsupervised Aspect-Sentiment Co-Extraction

Neural Machine Translation models are sensitive to noise in the input texts, such as misspelled words and ungrammatical constructions. Existing robustness techniques generally fail when faced with unseen types of noise and their performance degrades on clean texts. In this paper, we focus on three types of realistic noise that are commonly generated by humans and introduce the idea of visual context to improve translation robustness for noisy texts. In addition, we describe a novel error
correction training regime that can be used as an auxiliary task to further improve translation robustness. Experiments on EnglishFrench and English-German translation show that both multimodal and error correction components improve model robustness to noisy texts, while still retaining translation quality on clean texts.

Visual Cues and Error Correction for Translation Robustness

Existing sarcasm detection systems focus on exploiting linguistic markers, context, or user-level priors. However, social studies suggest that the relationship between the author and the audience can be equally relevant for the sarcasm usage and interpretation. In this work, we propose a framework jointly leveraging (1) a user context from their historical tweets together with (2) the social information from a user's conversational neighborhood in an interaction graph, to contextualize the interpretation of the post. We use graph attention networks (GAT) over users and tweets in a conversation thread, combined with dense user history representations. 
Apart from achieving state-of-the-art results on the recently published dataset of 19k Twitter users with 30K labeled tweets, adding 10M unlabeled tweets as context, our results indicate that the model contributes to interpreting the sarcastic intentions of an author more than to predicting the sarcasm perception by others.

Perceived and Intended Sarcasm Detection with Graph Attention Networks

In this paper, we present a new method for training a writing improvement model adapted to the writer’s first language (L1) that goes beyond grammatical error correction (GEC). Without using annotated training data, we rely solely on pre-trained language models finetuned with parallel corpora of reference translation aligned with machine translation. We evaluate our model with corpora of academic papers written in English by L1 Portuguese and L1 Spanish scholars and a reference corpus of expert academic English. We show that our model is able to address specific L1-influenced writing and more complex linguistic phenomena than existing methods, outperforming what a state-of-the-art GEC system can achieve in this regard. Our code and data are open to other researchers.

Raise a Child in Large Language Model: Towards Effective and Generalizable Fine-tuning

Downloads

Next from EMNLP 2021

A Simple and Effective Positional Encoding for Transformers

Similar lecture

The Low-Resource Double Bind: An Empirical Study of Pruning for Low-Resource Machine Translation

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Raise a Child in Large Language Model: Towards Effective and Generalizable Fine-tuning

.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 EMNLP 2021

A Simple and Effective Positional Encoding for Transformers

Similar lecture

The Low-Resource Double Bind: An Empirical Study of Pruning for Low-Resource Machine Translation

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Downloads