Papers by Yangfeng Ji
SideControl: Controlled Open-domain Dialogue Generation via Additive Side Networks (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods to generate pre-trained language models with attributes are expensive and overfitted on small training sets. |
| Approach: | They propose a novel approach to control the generation of Transformer-based pre-trained language models using a new control attributes loss framework. |
| Outcome: | The proposed method is shown to perform well with very limited training samples. |
Balanced Adversarial Training: Balancing Tradeoffs between Fickleness and Obstinacy in NLP Models (2022.emnlp-main)
Copied to clipboard
| Challenge: | Traditional adversarial examples involve finding a small perturbation that does not change an input’s true label but confuses the classifier into outputting a different prediction. |
| Approach: | They propose to use contrastive learning to increase model robustness against fickle adversarial examples by reducing the vulnerability of adversarials to fickle ones. |
| Outcome: | The proposed method improves model robustness against fickle and obstinate adversarial examples. |
An Empirical Comparison on Imitation Learning and Reinforcement Learning for Paraphrase Generation (D19-1)
Copied to clipboard
| Challenge: | Existing methods to generate paraphrases are not trivial and often fail in practice. |
| Approach: | They propose to use imitation learning to boost the performance of generating paraphrases by using a pointer-generator model. |
| Outcome: | The proposed model outperforms the state-of-the-art methods on the benchmark datasets. |
Contrastive Data and Learning for Natural Language Processing (2022.naacl-tutorials)
Copied to clipboard
| Challenge: | Current NLP models heavily rely on effective representation learning algorithms. |
| Approach: | This tutorial introduces contrastive learning and provides an introduction to the techniques. |
| Outcome: | This tutorial provides an introduction to the fundamentals of contrastive learning approaches and the theory behind them. |
Data Selection for Fine-tuning Large Language Models Using Transferred Shapley Values (2023.acl-srw)
Copied to clipboard
| Challenge: | Large language models (LMs) have been shown to be highly effective for identifying harmful training instances, but dataset size and model complexity constraints limit the ability to apply Shapley-based data valuation to fine-tuning large pre-trained language models. |
| Approach: | They propose an algorithm that aggregates Shapley values from subsets for valuation of entire training set and a value transfer method that leverages value information extracted from a simple classifier trained using representations from the target language model. |
| Outcome: | The proposed method outperforms existing methods on benchmark datasets and can filter fine-tuning data to increase language model performance compared to training with the full fine-uning dataset. |
Finding Friends and Flipping Frenemies: Automatic Paraphrase Dataset Augmentation Using Graph Theory (2020.findings-emnlp)
Copied to clipboard
| Challenge: | Having high quality annotated data is crucial for training supervised machine learning models. |
| Approach: | They propose automated methods to improve NLP datasets by viewing them as graphs with expected semantic properties. |
| Outcome: | The proposed methods improve paraphrase models on pre-trained datasets. |
Reevaluating Adversarial Examples in Natural Language (2020.findings-emnlp)
Copied to clipboard
| Challenge: | State-of-the-art adversarial examples lack a common definition of what constitutes success . human surveys show that to preserve semantics, we need to increase the minimum cosine similarities between the embeddings of swapped words and between the sentence encodings of original and perturbed sentences. |
| Approach: | They propose a unified definition of what constitutes a successful adversarial example . they propose four categories of constraints that are used to define adversarials . |
| Outcome: | The proposed framework is based on the outputs of two state-of-the-art synonym substitution attacks. |
White-box Testing of NLP models with Mask Neuron Coverage (2022.findings-naacl)
Copied to clipboard
| Challenge: | Recent research has shown that black-box testing is not applicable to NLP models. |
| Approach: | They propose a set of white-box testing methods that are customized for transformer-based NLP models and adapt them to a black-box test suite. |
| Outcome: | The proposed methods can reduce testing suites by 60% while retaining failing tests, thereby concentrating faultdetection power of the test suite. |
Unsupervised Concept Vector Extraction for Bias Control in LLMs (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large language models (LLMs) are known to perpetuate stereotypes and exhibit biases. |
| Approach: | They propose a method that extracts concept representations via probability weighting without labeled data and efficiently selects a steering vector for measuring and manipulating the model’s representation. |
| Outcome: | The proposed method can be used to predict gender bias and generalizes to racial bias. |
In-Context Learning (and Unlearning) of Length Biases (2025.naacl-long)
Copied to clipboard
| Challenge: | Existing work has demonstrated the ability of large language models to learn lexical and label biases in-context negatively impacts performance and robustness of models. |
| Approach: | They investigate the impact of length biases on in-context learning by analyzing model length information in-constext. |
| Outcome: | The proposed model learns length biases in the context window without parameter updates. |
HittER: Hierarchical Transformers for Knowledge Graph Embeddings (2021.emnlp-main)
Copied to clipboard
| Challenge: | Existing knowledge graph embedding methods to learn representations of knowledge graphs are conceptually simple and can be applied to tasks like factoid question answering (Saxena et al., 2020) and reasoning. |
| Approach: | They propose a Hierarchical Transformer model to jointly learn Entity-relation composition and Relational contextualization based on a source entity’s neighborhood. |
| Outcome: | The proposed model achieves state-of-the-art on multiple link prediction datasets and can be integrated into BERT and demonstrate its effectiveness on two Freebase factoid question answering datasets. |
Generating Hierarchical Explanations on Text Classification via Feature Interaction Detection (2020.acl-main)
Copied to clipboard
| Challenge: | Existing methods for generating explanations for neural networks ignore feature interactions between words and phrases. |
| Approach: | They propose to build hierarchical explanations by detecting feature interactions by combining words and phrases at different levels of the hierarchy. |
| Outcome: | The proposed method is evaluated on two benchmark datasets, via automatic and human evaluations. |
The Good, the Bad, and the Debatable: A Survey on the Impacts of Data for In-Context Learning (2025.emnlp-main)
Copied to clipboard
| Challenge: | In-context learning (ICL) is an emergent capability of large language models that allows them to learn new tasks at inference time without updating parameter updates. |
| Approach: | They propose to examine the relationship between data and in-context learning by examining the qualities of demonstrations that are desirable when selecting demonstrations, the "bad" qualities of demonstrators that can negatively impact the model and the "debatable" qualities. |
| Outcome: | The proposed model can learn unseen tasks by seeing a number of examples in the context window without updating parameters. |
Learning Variational Word Masks to Improve the Interpretability of Neural Text Classifiers (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for improving model interpretability require prior information or human annotations as additional inputs. |
| Approach: | They propose a variational word mask method to automatically learn task-specific important words and reduce irrelevant information on classification, which ultimately improves model interpretability. |
| Outcome: | The proposed method improves model prediction accuracy and interpretability on seven datasets. |
REV: Information-Theoretic Evaluation of Free-Text Rationales (2023.acl-long)
Copied to clipboard
| Challenge: | Existing metrics for rationale evaluation focus on the association between the rationale and a label, whereas REV is more sensitive to new information in free-text rationales. |
| Approach: | They propose a metric called REV to quantify the amount of new, label-relevant information in a rationale beyond the information already available in the input or the label. |
| Outcome: | The proposed metric is consistent with human judgments on rationale evaluations and provides more sensitive measurements of new information in free-text rationales. |
FlowEval: A Consensus-Based Dialogue Evaluation Framework Using Segment Act Flows (2022.emnlp-main)
Copied to clipboard
| Challenge: | Despite recent progress in dialogue evaluation, how to develop automatic metrics remains an open problem. |
| Approach: | They propose a consensus-based framework for dialog evaluation using segment act flows . they propose to crowdsource a large-scale dataset for it to be evaluated . |
| Outcome: | The proposed framework can reach the best or comparable correlation with human evaluation. |
Self-training with Two-phase Self-augmentation for Few-shot Dialogue Generation (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for self-training from meaning representations (MRs) are noisy or uninformative for the model to learn from. |
| Approach: | They propose a two-phase procedure to generate high-quality pseudo-labeled MR-to-Text pairs by aggregating multiple perturbed latent representations from each MR. |
| Outcome: | Empirical results on two benchmark datasets show that the proposed procedure outperforms existing methods on automatic and human evaluations. |
The Amazing World of Neural Language Generation (2020.emnlp-tutorials)
Copied to clipboard
| Challenge: | Recent years have seen a paradigm shift in neural text generation due to advances in deep contextual language modeling and transfer learning. |
| Approach: | They will discuss how and why NLG models succeed/fail at generating coherent text. |
| Outcome: | This paper will discuss how and why these models succeed/fail at generating coherent text, and provide insights on several applications. |
Addressing Both Statistical and Causal Gender Fairness in NLP Models (2024.findings-naacl)
Copied to clipboard
| Challenge: | Statistical fairness stipulates equivalent outcomes for all protected groups, whereas causal fairness prescribes that a model makes the same prediction for an individual regardless of their protected characteristics. |
| Approach: | They propose to use statistical and causal debiasing methods to reduce gender bias in NLP models. |
| Outcome: | The proposed methods reduce gender bias measured by the targeted metric, but not on other bias metrics. |
Pointwise Paraphrase Appraisal is Potentially Problematic (2020.acl-srw)
Copied to clipboard
| Challenge: | prevailing methods for paraphrase identification models are binary classification problems . current methods do not provide consistent and robust performance on unseen samples and real world problems. |
| Approach: | They propose to use binary classification to evaluate paraphrase identification models . they propose to improve methods for fine-tuning BERT models by pairing two sentences as one sequence . |
| Outcome: | The proposed methods may fail on simple tasks like identifying pairs with two identical sentences. |
Neural Text Generation in Stories Using Entity Representations as Context (N18-1)
Copied to clipboard
| Challenge: | Existing models of text generation that explicitly represent entities are based on the use of words and entities. |
| Approach: | They propose a neural model that explicitly represents entities mentioned in the text . they use vectors that are updated as the text proceeds to improve automatic evaluations . |
| Outcome: | The proposed model improves mention generation, sentence selection, and sentence generation. |
PLAtE: A Large-scale Dataset for List Page Web Extraction (2023.acl-industry)
Copied to clipboard
Aidan San, Yuan Zhuang, Jan Bakus, Colin Lockard, David Ciemiewicz, Sandeep Atluri, Kevin Small, Yangfeng Ji, Heba Elfardy
| Challenge: | Existing methods for web extraction are limited by the limited number of available large-scale datasets. |
| Approach: | They introduce a dataset that focuses on shopping data and a list page web extraction task. |
| Outcome: | The proposed dataset is the first large-scale list page web extraction dataset . it contains 52,898 items and 156,014 attributes, making it the first dataset based on this task . |
Explaining Neural Network Predictions on Sentence Pairs via Learning Word-Group Masks (2021.naacl-main)
Copied to clipboard
| Challenge: | Existing methods to explain neural network models are computationally inefficient for text inputs. |
| Approach: | They propose a method to implicitly detect word correlations by grouping correlated words from input text pairs together and measuring their contribution to corresponding NLP tasks. |
| Outcome: | The proposed method is evaluated with two different model architectures across four datasets. |
A Tale of Two Linkings: Dynamically Gating between Schema Linking and Structural Linking for Text-to-SQL Parsing (2020.coling-main)
Copied to clipboard
| Challenge: | Existing methods for text-to-SQL semantic parsing require strict structured prediction due to its application scenario where the output SQL will be sent to an executor program directly. |
| Approach: | They propose to use schema linking and structural linking to link NL to the database schema. |
| Outcome: | The proposed method shows significant gains on the Spider dataset. |