Peer review is a core component of scholarly publishing, yet it is time-consuming, requires considerable expertise, and is prone to error. The applications of NLP for peer reviewing assistance aim to mitigate those issues, but the lack of clearly licensed datasets and multi-domain corpora prevent the systematic study of NLP for peer review. To remedy this, we introduce NLPeer -- the first ethically sourced multidomain corpus of more than 5k papers and 11k review reports from five different venues. In addition to the new datasets of paper drafts, camera-ready versions and peer reviews from the NLP community, we establish a unified data representation, and augment previous peer review datasets to include parsed, structured paper representations, rich metadata and versioning information. Our work paves the path towards systematic, multi-faceted, evidence-based study of peer review in NLP and beyond. We make NLPeer publicly available.
Natural language processing researchers develop models of grammar, meaning and human communication based on written text. Due to task and data differences, what is considered text can vary substantially across studies. A conceptual framework for systematically capturing these differences is lacking. We argue that clarity on the notion of text is crucial for reproducible and generalizable NLP. Towards that goal, we propose common terminology to discuss the production and transformation of textual data, and introduce a two-tier taxonomy of linguistic and non-linguistic elements that are available in textual sources and can be used in NLP modeling. We apply this taxonomy to survey existing work that extends the notion of text beyond the conservative language-centered view. We outline key desiderata and challenges of the emerging inclusive approach to text in NLP, and suggest systematic community-level reporting as a crucial next step to consolidate the discussion.
Stance detection deals with the identification of an author's stance towards a target and is applied on various text domains like social media and news. In many cases, inferring the stance is challenging due to insufficient access to contextual information. Complementary context can be found in knowledge bases but integrating the context into pretrained language models is non-trivial due to their graph structure. In contrast, we explore an approach to integrate contextual information as text which aligns better with transformer architectures. Specifically, we train a model consisting of dual encoders which exchange information via cross-attention. This architecture allows for integrating contextual information from heterogeneous sources. We evaluate context extracted from structured knowledge sources and from prompting large language models. Our approach is able to outperform competitive baselines (1.9pp on average) on a large and diverse stance detection benchmark, both (1) in-domain, i.e. for seen targets, and (2) out-of-domain, i.e. for targets unseen during training. Our analysis shows that it is able to regularize for spurious label correlations with target-specific cue words.
We compare sequential fine-tuning with a model for multi-task learning in the context where we are interested in boosting performance on two tasks, one of which depends on the other. We test these models on the FigLang2022 shared task which requires participants to predict language inference labels on figurative language along with corresponding textual explanations of the inference predictions. Our results show that while sequential multi-task learning can be tuned to be good at the first of two target tasks, it performs less well on the second and additionally struggles with overfitting. Our findings show that simple sequential fine-tuning of text-to-text models is an extraordinarily powerful method for cross-task knowledge transfer while simultaneously predicting multiple interdependent targets. So much so, that our best model achieved the (tied) highest score on the task.
Misinformation emerges in times of uncertainty when credible information is limited. This is challenging for NLP-based fact-checking as it relies on counter-evidence, which may not yet be available. Despite increasing interest in automatic fact-checking, it is still unclear if automated approaches can realistically refute harmful real-world misinformation. Here, we contrast and compare NLP fact-checking with how professional fact-checkers combat misinformation in the absence of counter-evidence. In our analysis, we show that, by design, existing NLP task definitions for fact-checking cannot refute misinformation as professional fact-checkers do for the majority of claims. We then define two requirements that the evidence in datasets must fulfill for realistic fact-checking: It must be (1) sufficient to refute the claim and (2) not leaked from existing fact-checking articles. We survey existing fact-checking datasets and find that all of them fail to satisfy both criteria. Finally, we perform experiments to demonstrate that models trained on a large-scale fact-checking dataset rely on leaked evidence, which makes them unsuitable in real-world scenarios. Taken together, we show that current NLP fact-checking cannot realistically combat real-world misinformation because it depends on unrealistic assumptions about counter-evidence in the data.
Pairing a lexical retriever with a neural re-ranking model has set state-of-the-art performance on large-scale information retrieval datasets. This pipeline covers scenarios like question answering or navigational queries, however, for information-seeking scenarios, users often provide information on whether a document is relevant to their query in form of clicks or explicit feedback. Therefore, in this work, we explore how relevance feedback can be directly integrated into neural re-ranking models by adopting few-shot and parameter-efficient learning techniques. Specifically, we introduce a kNN approach that re-ranks documents based on their similarity with the query and the documents the user considers relevant. Further, we explore Cross-Encoder models that we pre-train using meta-learning and subsequently fine-tune for each query, training only on the feedback documents. To evaluate our different integration strategies, we transform four existing information retrieval datasets into the relevance feedback scenario. Extensive experiments demonstrate that integrating relevance feedback directly in neural re-ranking models improves their performance, and fusing lexical ranking with our best performing neural re-ranker outperforms all other methods by 5.2 nDCG@20.
Few-shot Intent Classification (FSIC) is one of the key challenges in modular task-oriented dialog systems. While advanced FSIC methods are similar in using pretrained language models to encode texts and nearest neighbour-based inference for classification, these methods differ in details. They start from different pretrained text encoders, use different encoding architectures with varying similarity functions, and adopt different training regimes. Coupling these mostly independent design decisions and the lack of accompanying ablation studies are big obstacle to identify the factors that drive the reported FSIC performance. We study these details across three key dimensions: (1) Encoding architectures: Cross-Encoder vs Bi-Encoders; (2) Similarity function: Parameterized (i.e., trainable) functions vs non-parameterized function; (3) Training regimes: Episodic meta-learning vs the straightforward (i.e., non-episodic) training. Our experimental results on seven FSIC benchmarks reveal three important findings. First, the unexplored combination of the cross-encoder architecture (with parameterized similarity scoring function) and episodic meta-learning consistently yields the best FSIC performance. Second, Episodic training yields a more robust FSIC classifier than non-episodic one. Third, in meta-learning methods, splitting an episode to support and query sets is not a must. Our findings paves the way for conducting state-of-the-art research in FSIC and more importantly raise the community's attention to details of FSIC methods. We release our code and data publicly.
Current multimodal models, aimed at solving Vision and Language (V+L) tasks, predominantly repurpose Vision Encoders (VE) as feature extractors. While many VEs -- of different architectures, trained on different data and objectives -- are publicly available, they are not designed for the downstream V+L tasks. Nonetheless, most current work assumes that a \textit{single} pre-trained VE can serve as a general-purpose encoder. In this work, we evaluate whether the information stored within different VEs is complementary, i.e. if providing the model with features from multiple VEs can improve the performance on a target task. We exhaustively experiment with three popular VEs on six downstream V+L tasks and analyze the attention and VE-dropout patterns. Our results and analyses suggest that diverse VEs complement each other, resulting in improved downstream V+L task performance, where the improvements are not due to simple ensemble effects (i.e. the performance does not always improve when increasing the number of encoders). We demonstrate that future VEs, which are not \textit{repurposed}, but explicitly \textit{designed} for V+L tasks, have the potential of improving performance on the target V+L tasks.
Activation functions can have a significant impact on reducing the topological complexity of input data and therefore improve the performance of the model. Selecting a suitable activation function is an essential step in neural model design. However, the choice of activation function is seldom discussed or explored in Transformer-based language models. Their activation functions are chosen beforehand and then remain fixed from pre-training to fine-tuning. As a result, the inductive biases they imposed on models cannot be adjusted during this long life cycle. Moreover, subsequently developed models (e.g., RoBERTa, BART, and GPT-3) often follow up prior work (e.g., BERT) to use the same activation function without justification. In this paper, we investigate the effectiveness of using Rational Activation Function (RAF), a learnable activation function, in the Transformer architecture. In contrast to conventional, predefined activation functions, RAFs can adaptively learn optimal activation functions during training according to input data. Our experiments show the RAF-based Transformer (RAFT) achieves a lower validation perplexity than a vanilla BERT with the GELU function. We further evaluate RAFT on downstream tasks in low- and full-data settings. Our results show that RAFT outperforms the counterpart model across the majority of tasks and settings. For instance, RAFT outperforms vanilla BERT on the GLUE benchmark by 5.71 points on average in low-data scenario (where 100 training examples are available) and by 2.05 points on SQuAD in full-data setting. Analysis of the shapes of learned RAFs further unveils that they substantially vary between different layers of the pre-trained model and mostly look very different from conventional activation functions. RAFT opens a new research direction for analyzing and interpreting pre-trained models according to the learned activation functions.
Question Answering (QA) systems are increasingly deployed in applications where they support real-world decisions. However, state-of-the-art models rely on deep neural networks, which are difficult to interpret by humans. Inherently interpretable models or post hoc explainability methods can help users to comprehend how a model arrives at its prediction and, if successful, increase their trust in the system. Furthermore, researchers can leverage these insights to develop new methods that are more accurate and less biased. In this paper, we introduce SQuARE v2, the new version of SQuARE, to provide an explainability infrastructure for comparing models based on methods such as saliency maps and graph-based explanations. While saliency maps are useful to inspect the importance of each input token for the model's prediction, graph-based explanations from external Knowledge Graphs enable the users to verify the reasoning behind the model prediction. In addition, we provide multiple adversarial attacks to compare the robustness of QA models. With these explainability methods and adversarial attacks, we aim to ease the research on trustworthy QA models. SQuARE is available on https://square.ukp-lab.de.