LOSS FUNCTION

CLEAR: Cross-Lingual Enhancement in Retrieval via Reverse-training (2026.acl-long)

Challenge:	Large Language Models (LLMs) have been shown to be effective in complex language tasks, but their potential to perpetuate biases poses significant concerns.
Approach:	They propose a new framework employing Direct Preference Optimization to mitigate biases in LLMs.
Outcome:	The proposed model outperforms the baseline model on almost all bias benchmarks and achieves better performance than open-source models.

Copied to clipboard

Seungyoon Lee, Minhyuk Kim, Seongtae Hong, Youngjoon Jang, Dongsuk Oh, Heuiseok Lim

Reference Free Domain Adaptation for Translation of Noisy Questions with Question Specific Rewards (2023.findings-emnlp)

Challenge:	Existing multilingual embedding models often struggle to capture cross-lingual alignment during training.
Approach:	They propose a novel loss function that leverages an English passage as a bridge to strengthen alignments between target language and English.
Outcome:	The proposed model improves retrieval performance across cross-lingual scenarios while minimizing performance degradation in English.

Copied to clipboard

Baban Gain, Ramakrishna Appicharla, Soumya Chennabasavaraj, Nikesh Garera, Asif Ekbal, Muthusamy Chelliah

Transformer-based Approach for Predicting Chemical Compound Structures (2020.aacl-main)

Challenge:	Creating a synthetic parallel corpus from noisy data is also difficult due to its noisy nature.
Approach:	They propose a training methodology that fine-tunes the NMT system only using source-side data to balance adequacy and fluency.
Outcome:	The proposed method surpasses the MLE-based fine-tuning approach by achieving a 1.9 BLEU improvement.

Copied to clipboard

Yutaro Omote, Kyoumoto Matsushita, Tomoya Iwakura, Akihiro Tamura, Takashi Ninomiya

Train One Get One Free: Partially Supervised Neural Network for Bug Report Duplicate Detection and Clustering (N19-2)

Challenge:	Existing methods to predict chemical compound structures from their names are limited and use handcrafted rules.
Approach:	They propose a Transformer-based model that predicts SMILES strings from chemical compound names instead of handcrafted rules.
Outcome:	The proposed model achieves higher F-measures than the existing model and the existing one.

Copied to clipboard

Lahari Poddar, Leonardo Neves, William Brendel, Luis Marujo, Sergey Tulyakov, Pradeep Karuturi

Unsupervised KB-to-Text Generation with Auxiliary Triple Extraction using Dual Learning (2020.aacl-main)

Challenge:	Existing methods for duplicate classification require manual review and assigning bugs to the correct teams.
Approach:	They propose a loss function that can detect duplicate bug reports and aggregate them into latent topics without supervision.
Outcome:	The proposed model outperforms state-of-the-art methods for duplicate classification on both cases and can learn meaningful latent clusters without supervision.

Copied to clipboard

Zihao Fu, Bei Shi, Lidong Bing, Wai Lam

Low-Resource NLP

Challenge:	Existing methods to generate text from KB triples are limited and expensive . a novel approach is proposed to train the generation model in unsupervised way .
Approach:	They propose a method which trains the generation model in a completely unsupervised way with unaligned raw text data and KB triples.
Outcome:	The proposed method outperforms existing methods and is cost-effective.

Simple Question Answering with Subgraph Ranking and Joint-Scoring (N19-1)

Copied to clipboard

Wenbo Zhao, Tagyoung Chung, Anuj Goyal, Angeliki Metallinou

Multimodality

Challenge:	Knowledge graph based simple question answering is a major area of research in question answering.
Approach:	They propose a framework to describe and analyze existing knowledge graph based simple question answering approaches.
Outcome:	The proposed model achieves a state-of-the-art (85.44% accuracy) on the SimpleQuestions dataset.

Neural OCR Post-Hoc Correction of Historical Corpora (2021.tacl-1)

Copied to clipboard

Lijun Lyu, Maria Koutraki, Martin Krickl, Besnik Fetahu

Visual Data in NLP

Challenge:	Optical character recognition (OCR) is crucial for a deeper access to historical collections.
Approach:	They propose a neural approach based on a combination of recurrent (RNN) and deep convolutional network (ConvNet) to correct OCR transcription errors.
Outcome:	The proposed model reduces the word error rate of 32.3% by more than 89% on a historical book corpus in German language.

Bag-of-Vectors Autoencoders for Unsupervised Conditional Text Generation (2022.aacl-main)

Copied to clipboard

Florian Mai, James Henderson

Personalized Filled-pause Generation with Group-wise Prediction Models (2022.lrec-1)

Challenge:	Existing methods to learn mappings in the embedding space of text autoencoders are limited to a single-vector embeddment, which limits how much information can be retained.
Approach:	They propose a method to learn mappings in the embedding space of an autoencoder by extending it to Bag-of-Vectors Autoencodeurs (BoV-AEs) this allows to encode and reconstruct much longer texts than standard autoencodings .
Outcome:	The proposed method performs better than a standard autoencoder on unsupervised sentiment transfer.

Copied to clipboard

Yuta Matsunaga, Takaaki Saeki, Shinnosuke Takamichi, Hiroshi Saruwatari

Representation Learning

Challenge:	Disfluency generation is a method to generate personalized filled pauses (FPs) compared with fluent text generation, it is difficult to predict them because of the sparsity of position and frequency difference between more and less frequently used FPs.
Approach:	They propose a method to generate personalized filled pauses (FPs) by group-wise prediction models.
Outcome:	The proposed method generates personalized filled pauses (FPs) with group-wise prediction models.

Capturing the Relationship Between Sentence Triplets for LLM and Human-Generated Texts to Enhance Sentence Embeddings (2024.findings-eacl)

Copied to clipboard

Na Min An, Sania Waheed, James Thorne

Representation Learning

Challenge:	Recent advances in building sentence embedding models have centered on replacing traditional human-generated text datasets with those generated by LLMs.
Approach:	They propose a loss function that incorporates Positive-Negative sample Augmentation within the contrastive learning objective to enhance sentence embeddings using both human and LLM-generated datasets.
Outcome:	The proposed model mitigates the sentence anisotropy problem in Wikipedia corpus and improves Spearman’s correlation in standard Semantic Textual Similarity (STS) tasks (+1.47% compared to CLHAIF).

CoRet: Improved Retriever for Code Editing (2025.acl-short)

Copied to clipboard

Fabio James Fehr, Prabhu Teja S, Luca Franceschi, Giovanni Zappella

Information Retrieval

Challenge:	Existing encoder models perform poorly in repository-level retrieval for code-editing tasks.
Approach:	They propose a loss function for code retrieval that integrates code semantics, repository structure, and call-graph dependencies.
Outcome:	The proposed model significantly improves retrieval recall by at least 15 percentage points over existing models on SWE-bench and Long Code Arena’s bug localisation datasets.

ReAugKD: Retrieval-Augmented Knowledge Distillation For Pre-trained Language Models (2023.acl-short)

Copied to clipboard

Jianyi Zhang, Aashiq Muhamed, Aditya Anantharaman, Guoyin Wang, Changyou Chen, Kai Zhong, Qingjun Cui, Yi Xu, Belinda Zeng, Trishul Chilimbi, Yiran Chen

Towards Efficient and Multifaceted Computer-assisted Pronunciation Training Leveraging Hierarchical Selective State Space Model and Decoupled Cross-entropy Loss (2025.naacl-long)

Challenge:	Knowledge distillation (KD) is an effective compression technique to derive a smaller student model from a larger teacher model by transferring the knowledge embedded in the teacher's network.
Approach:	They propose a framework and loss function that preserves the semantic similarities of teacher and student training examples to enable the student to retrieve from the knowledge base effectively.
Outcome:	The proposed framework preserves the semantic similarities of teacher and student training examples to achieve state-of-the-art performance on the GLUE benchmark.

Copied to clipboard

Fu-An Chao, Berlin Chen

Green & Sustainable NLP

Challenge:	APA and MDD are two of the main tasks of computer-assisted pronunciation training (CAPT) systems.
Approach:	They propose a computer-assisted pronunciation training approach that integrates APA and MDD tasks in parallel.
Outcome:	The proposed approach improves on APA and MDD tasks, and achieves an F1 score of 63.85%.

Towards Robust Extractive Question Answering Models: Rethinking the Training Methodology (2024.findings-emnlp)

Copied to clipboard

Son Tran, Matt Kretchmar

Robustness in NLP

Challenge:	Existing models lack robustness against distribution shifts and adversarial attacks when training on unanswerable questions in EQA datasets.
Approach:	They propose a novel loss function for the EQA problem to improve the robustness of extractive question answering models by adding adversarial questions to a crowdsourcing process.
Outcome:	The proposed method maintains in-domain performance while improving on out-of-domain datasets.

LRQuant: Learnable and Robust Post-Training Quantization for Large Language Models (2024.acl-long)

Copied to clipboard

Jiaqi Zhao, Miao Zhang, Chao Zeng, Ming Wang, Xuebo Liu, Liqiang Nie

Robustness in NLP

Challenge:	Existing methods for post-training quantization (PTQ) are limited by the complexity of the quantization parameter and performance degradations when tested on unseen datasets.
Approach:	They propose a learnable smooth-based PTQ framework that allows for rapid adaptation during testing.
Outcome:	The proposed framework improves performance on unseen datasets and reduces memory constraints.

Bring More Attention to Syntactic Symmetry for Automatic Postediting of High-Quality Machine Translations (2023.acl-short)

Copied to clipboard

Baikjin Jung, Myungji Lee, Jong-Hyeok Lee, Yunsu Kim

Machine Translation

Challenge:	Existing APE systems are not good at handling high-quality MTs even for a language pair with abundant data resources, English–German.
Approach:	They propose a linguistically motivated method of regularization that encourages symmetric self-attention on the given MT.
Outcome:	The proposed method improves the state-of-the-art architecture’s APE quality for high-quality MTs.

Implicit Unlikelihood Training: Improving Neural Text Generation with Reinforcement Learning (2021.eacl-main)

Copied to clipboard

Evgeny Lagutin, Daniil Gavrilov, Pavel Kalaidin

A Study of Syntactic Multi-Modality in Non-Autoregressive Machine Translation (2022.naacl-main)

Challenge:	Existing approaches to language modeling use autoregressive methods, but they can produce repetitive results.
Approach:	They propose to add a loss function for regularization to avoid unwanted properties, such as contradiction or repetition, to a language model by using policy gradient reinforcement learning.
Outcome:	The proposed method reduces repetition without impacting the language model quality.

Copied to clipboard

Kexun Zhang, Rui Wang, Xu Tan, Junliang Guo, Yi Ren, Tao Qin, Tie-Yan Liu

SpanEmo: Casting Multi-label Emotion Classification as Span-prediction (2021.eacl-main)

Challenge:	Non-autoregressive translation models suffer from the multi-modality problem when a source sentence corresponds to multiple correct translations.
Approach:	They propose to decompose the syntactic multi-modality problem into short- and long-range models and evaluate them on synthesized and real datasets.
Outcome:	The proposed loss functions can handle short- and long-range syntactic multi-modalities better than existing models.

Copied to clipboard

Hassan Alhuzali, Sophia Ananiadou

Information Retrieval

Challenge:	Current approaches to ER ignore potential ambiguities, in which multiple emotions overlap.
Approach:	They propose a model "SpanEmo" which casts multi-label emotion classification as span-prediction and introduces a loss function focused on modelling multiple co-existing emotions in a sentence.
Outcome:	The proposed model can predict multiple co-existing emotions in a sentence and improve model performance and learning meaningful associations between labels and words in the sentence.

Direct Multi-Turn Preference Optimization for Language Agents (2024.emnlp-main)

Copied to clipboard

Wentao Shi, Mengqi Yuan, Junkang Wu, Qifan Wang, Fuli Feng

Challenge:	Extensive experiments on three multi-turn agent task datasets confirm the effectiveness and superiority of the DMPO loss function.
Approach:	They propose a novel loss function for multi-turn agent tasks that replaces the policy constraint with the state-action occupancy measure constraint and adds length normalization to the Bradley-Terry model.
Outcome:	Experiments on three multi-turn agent task datasets confirm the effectiveness and superiority of the proposed loss function.

Improving End-to-End Bangla Speech Recognition with Semi-supervised Training (2020.findings-emnlp)

Copied to clipboard

Nafis Sadeq, Nafis Tahmid Chowdhury, Farhan Tanvir Utshaw, Shafayat Ahmed, Muhammad Abdullah Adnan

Multimodality

Challenge:	Existing methods to train speech recognition systems require large annotated corpus.
Approach:	They propose a semi-supervised training approach that exploits large unpaired audio and text data to improve the performance of an automatic speech recognition system.
Outcome:	The proposed method reduces the WER of the system from 37% to 31.9%.

Multimodal Document-level Triple Extraction via Dynamic Graph Enhancement and Relation-Aware Reflection (2025.findings-emnlp)

Copied to clipboard

Xiang Li, Runhai Jiao, Zhou Changyu, Shoupeng Qiao, Ruojiao Qiao, Ruifan Li

Multimodality

Challenge:	Existing methods for extracting structured triples knowledge from multimodal documents face limitations in simultaneously processing long textual content and multiple associated images for triple extraction.
Approach:	They propose a multimodal document-level triple extraction framework that integrates multimodal text and visual content into a large language model and injects the global information and external knowledge into the model.
Outcome:	The proposed framework outperforms the state-of-the-art methods and fills the gap in multimodal document extraction.

Influences on LLM Calibration: A Study of Response Agreement, Loss Functions, and Prompt Styles (2025.acl-long)

Copied to clipboard

Yuxi Xia, Pedro Henrique Luz De Araujo, Klim Zaporojets, Benjamin Roth

Towards Integration of Discriminability and Robustness for Document-Level Relation Extraction (2023.eacl-main)

Challenge:	Existing studies neglect to measure the generalization of their methods to other prompt styles and different sizes of LLMs.
Approach:	They propose a framework that trains an auxiliary model for confidence estimation that aggregates responses from multiple LLMs to capture inter-model agreement.
Outcome:	The proposed framework integrates response agreement and focal loss with binary cross-entropy to improve calibration from baselines.

Copied to clipboard

Jia Guo, Stanley Kok, Lidong Bing

Context-aware Entity Typing in Knowledge Graphs (2021.findings-emnlp)

Challenge:	Document-level relation extraction (DocRE) predicts relations for entity pairs relying on context-dependent reasoning . a large number of annotation errors can make it difficult to distinguish large semantically close relations .
Approach:	They propose a loss function to improve discriminability and robustness for DocRE . they also propose supervised contrastive learning and negative label sampling strategy .
Outcome:	The proposed method achieves state-of-the-art results on the DocRED dataset and its recently cleaned version.

Copied to clipboard

Weiran Pan, Wei Wei, Xian-Ling Mao

Knowledge Representation

Challenge:	Existing methods for knowledge graph entity typing are embedding-based and graph convolutional networks (GCNs) . Existing approaches for knowledge Graph Entity Typing (KGET) are incomplete and require multiple inference mechanisms.
Approach:	They propose a method that uses entities’ contextual information to infer missing types in knowledge graphs by using two inference mechanisms: N2T and Agg2T.
Outcome:	The proposed method can infer entities' missing types by completing two real-world KGs.

Document-Level Event Argument Extraction by Leveraging Redundant Information and Closed Boundary Loss (2022.naacl-main)

Copied to clipboard

Hanzhang Zhou, Kezhi Mao

Learning Semantic Sentence Embeddings using Sequential Pair-wise Discriminator (C18-1)

Challenge:	Document-level event argument extraction is a crucial subtask of event extraction.
Approach:	They propose to use redundant event information to extract multiple arguments from a document . they propose a loss function to classify Universum class by their open decision boundary .
Outcome:	The proposed model outperforms the previous state-of-the-art models by 3.35% in F1-score.

Copied to clipboard

Badri Narayana Patro, Vinod Kumar Kurmi, Sandeep Kumar, Vinay Namboodiri

Exploring the Trade-off Between Model Performance and Explanation Plausibility of Text Classifiers Using Human Rationales (2024.findings-naacl)

Challenge:	a novel method for obtaining sentence-level embeddings is proposed . the problem of obtaining a semantic embeddable sentence is at the core of understanding languages .
Approach:	They propose a method for obtaining sentence-level embeddings by using a sequential encoder-decoder framework.
Outcome:	The proposed method outperforms the state-of-the-art on a sentiment analysis task.

Copied to clipboard

Lucas Resck, Marcos M. Raimundo, Jorge Poco

Explainability & Interpretability in NLP

Challenge:	Saliency post-hoc explainability methods are important tools for understanding complex NLP models, but they may not align with human intuition, making the explanations not plausible.
Approach:	They propose a method for incorporating rationales into text classification models by augmenting the standard cross-entropy loss with a novel loss function inspired by contrastive learning.
Outcome:	The proposed approach enhances the plausibility of post-hoc explanations while preserving their faithfulness.

One Reference Is Not Enough: Diverse Distillation with Reference Selection for Non-Autoregressive Translation (2022.naacl-main)

Copied to clipboard

Chenze Shao, Xuanfu Wu, Yang Feng

Clickbait? Sensational Headline Generation with Auto-tuned Reinforcement Learning (D19-1)

Challenge:	Existing non-autoregressive neural machine translation models suffer from multimodality problem . multi-modality is not solved by a teacher forcing algorithm, limiting model capability .
Approach:	They propose a method that generates multiple reference translations for each source sentence . they compare the NAT output with all references and select the one that best fits the simulated model .
Outcome:	The proposed method achieves 29.82 BLEU with only one decoding pass on WMT14 En-De .

Copied to clipboard

Peng Xu, Chien-Sheng Wu, Andrea Madotto, Pascale Fung

Text Generation

Challenge:	Conventional abstractive headline generation methods do not optimize for maximum reader attention.
Approach:	They propose a model that generates sensational headlines without labeled data by classifying online headlines with many comments against a summarization model.
Outcome:	The proposed model generates sensational headlines without labeled data.

Keeping Consistency of Sentence Generation and Document Classification with Multi-Task Learning (D19-1)

Copied to clipboard

Toru Nishino, Shotaro Misawa, Ryuji Kano, Tomoki Taniguchi, Yasuhide Miura, Tomoko Ohkuma

Complexity-Weighted Loss and Diverse Reranking for Sentence Simplification (N19-1)

Challenge:	Existing automated generation of articles' characteristics is inconsistent if they are generated individually.
Approach:	They propose a multi-task learning model with a shared encoder and multiple decoders for each task.
Outcome:	The proposed model generates more consistent headlines, key phrases and categories . it outperforms baseline model on the ROUGE scores and generates fluent headlines .

Copied to clipboard

Reno Kriz, João Sedoc, Marianna Apidianaki, Carolina Zheng, Gaurav Kumar, Eleni Miltsakaki, Chris Callison-Burch

Cross-Lingual Document Retrieval with Smooth Learning (2020.coling-main)

Challenge:	Recent research has applied sequence-to-sequence (Seq2Sequen) models to text simplification . generic models tend to copy directly from the original sentence, resulting in outputs that are long and complex.
Approach:	They propose to incorporate word complexities into the loss function during training and generate a large set of diverse candidate simplifications at test time.
Outcome:	The proposed model can perform competitively with state-of-the-art systems while generating simpler sentences.

Copied to clipboard

Jiapeng Liu, Xiao Zhang, Dan Goldwasser, Xiao Wang

Information Retrieval

Challenge:	Cross-lingual document search is an information retrieval task in which the queries’ language and the documents’ language are different.
Approach:	They propose a robust framework that measures the relevance and a loss function that is a novel objective function.
Outcome:	The proposed framework achieves significant gains under commonly used ranking metrics on cross-lingual document retrieval task in a variety of languages.

OTSeq2Set: An Optimal Transport Enhanced Sequence-to-Set Model for Extreme Multi-label Text Classification (2022.emnlp-main)

Copied to clipboard

Jie Cao, Yin Zhang

Text Classification

Challenge:	Extreme multi-label text classification (XMTC) is a task of finding the most relevant subset labels from an extremely large label set.
Approach:	They propose an autoregressive sequence-to-set model for XMTC tasks named OTSeq2Set.
Outcome:	The proposed model outperforms other competitive baselines on 4 benchmark datasets.

Mix-of-Granularity: Optimize the Chunking Granularity for Retrieval-Augmented Generation (2025.coling-main)

Copied to clipboard

Zijie Zhong, Hanwen Liu, Xiaoya Cui, Xiaofan Zhang, Zengchang Qin

Syntactic Text Processing

Challenge:	Retrieval-augmented generation systems often use a fixed strategy to extract information from multiple sources.
Approach:	They propose a method that dynamically determines optimal granularity of a knowledge source based on input queries using a router.
Outcome:	The proposed method predicts optimal granularity levels and significantly improves performance in downstream tasks.

A Simple Log-based Loss Function for Ordinal Text Classification (2022.coling-1)

Copied to clipboard

François Castagnos, Martin Mihelich, Charles Dognin

Named Entity Recognition via Noise Aware Training Mechanism with Data Filter (2021.findings-acl)

Challenge:	Existing methods for ordinal text classification do not incorporate ordinal character into their feedback.
Approach:	They propose a new ordinal log-loss loss function that incorporates ordinal character into its feedback.
Outcome:	The proposed loss function outperforms state-of-the-art methods on four benchmark text classification datasets.

Copied to clipboard

Xiusheng Huang, Yubo Chen, Shun Wu, Jun Zhao, Yuantao Xie, Weijian Sun

On Eliciting Syntax from Language Models via Hashing (2024.emnlp-main)

Challenge:	Existing methods for named entity recognition (NER) do not distinguish noisy from hard samples.
Approach:	They propose a noise-aware-with-filter method to help model identify noisy samples . they propose 'incomplete trust' loss function which boosts L CRF with a robust term .
Outcome:	The proposed method outperforms the existing methods on six real-world Chinese and English NER datasets.

Copied to clipboard

Yiran Wang, Masao Utiyama

Syntactic Text Processing

Challenge:	Unsupervised parsing aims to infer syntactic structure from raw text . despite its importance, advancements in this task have been slow .
Approach:	They propose to use unsupervised parsing to infer syntactic structure from raw text . they upgrade the bit-level CKY to first-order to encode lexicon and syntax .
Outcome:	The proposed method shows competitive performance on various datasets.

How to Ask Good Questions? Try to Leverage Paraphrases (2020.acl-main)

Copied to clipboard

Xin Jia, Wenjie Zhou, Xu Sun, Yunfang Wu

Text Generation

Challenge:	Existing methods to generate human-like questions rely on paraphrases to generate good questions.
Approach:	They propose to integrate paraphrase knowledge into question generation to generate human-like questions by combining paraphrases with a back-translation method.
Outcome:	The proposed model achieves obvious performance gain over several strong baselines and human evaluation validates that it can ask questions of high quality by leveraging paraphrase knowledge.

EpLSA: Synergy of Expert-prefix Mixtures and Task-Oriented Latent Space Adaptation for Diverse Generative Reasoning (2024.lrec-main)

Copied to clipboard

Fujun Zhang, Xiangdong Su, Jiang Li, Rong Yan, Guanglai Gao

Reasoning

Challenge:	Existing models for diverse generative reasoning struggle to generate multiple unique and plausible results.
Approach:	They propose a model based on expert-prefix mixtures and task-oriented latent space adaptation for diverse generative reasoning.
Outcome:	The proposed model outperforms existing models on three types of generative reasoning tasks.

Learning to Substitute Words with Model-based Score Ranking (2025.naacl-long)

Copied to clipboard

Hongye Liu, Ricardo Henao

Visual Coherence Loss for Coherent and Visually Grounded Story Generation (2023.findings-acl)

Challenge:	Experimental results show that the proposed approach outperforms both masked language models and large language models.
Approach:	They propose a model-based scoring approach to quantify sentence quality . they propose 'loss function' that optimizes alignment between model predictions and sentence scores .
Outcome:	The proposed approach outperforms masked language models and large language models in the quantitative analysis of word substitutions.

Copied to clipboard

Xudong Hong, Vera Demberg, Asad Sayeed, Qiankun Zheng, Bernt Schiele

Visual Data in NLP

Challenge:	Existing visual storytelling models fail to generate correct referring expressions for characters, causing 60% of the generated stories to be lacking local coherence.
Approach:	They propose a loss function inspired by a linguistic theory of coherence for self-supervised learning for image sequence representations and a feature matching metric to check whether the models generate referring expressions correctly for characters in input image sequences.
Outcome:	The proposed features and loss function are effective for generating more coherent and visually grounded stories.

From Graph to Word Bag: Introducing Domain Knowledge to Confusing Charge Prediction (2024.lrec-main)

Copied to clipboard

Ang Li, Qiangchao Chen, Yiquan Wu, Xiang Zhou, Kun Kuang, Fei Wu, Ming Cai

Structured Data in NLP

Challenge:	Existing charge prediction methods have shown impressive performance, but they face significant challenges when dealing with confusing charges, such as Snatch and Robbery.
Approach:	They propose a novel approach which introduces domain knowledge regarding constituent elements to guide the model in making judgments on confusing charges, much like a judge’s reasoning process.
Outcome:	The proposed approach maintains exceptional performance in imbalanced label distributions.

HS-GC: Holistic Semantic Embedding and Global Contrast for Effective Text Clustering (2024.lrec-main)

Copied to clipboard

Chen Yang, Bin Cao, Jing Fan

Text Clustering

Challenge:	In this paper, we introduce Holistic Semantic Embedding and Global Contrast (HS-GC) to learn the instance- and cluster-level representations.
Approach:	They propose a novel loss function that exploits different layers of semantic information in a deep neural network to provide a more holistic semantic text representation.
Outcome:	The proposed model outperforms the state-of-the-art model on five text datasets and improves clustering accuracy of 5.9% and 3.2% on the StackOverflow and TREC datasets.

Language Concept Erasure for Language-invariant Dense Retrieval (2024.emnlp-main)

Copied to clipboard

Zhiqi Huang, Puxuan Yu, Shauli Ravfogel, James Allan

Mitigating Framing Bias with Polarity Minimization Loss (2023.findings-emnlp)

Challenge:	Multilingual models aim for language-invariant representations but still encode language identity.
Approach:	They propose a multi-task learning framework that induces language invariance in multilingual retrieval by reducing language-specific signals in the embedding space.
Outcome:	The proposed learning framework improves language-invariant dense retrieval over baselines on English retrieval data and general multilingual corpora.

Copied to clipboard

Yejin Bang, Nayeon Lee, Pascale Fung

Challenge:	polarity is a pervasive problem in modern media, misleading the understanding of what really happened via a skewed selection of information and language.
Approach:	They propose a loss function that encourages the model to minimize the polarity difference between the skewed input articles to reduce framing bias.
Outcome:	The proposed loss improves the model's ability to map polarity ends bidirectionally.

Rethinking Negative Pairs in Code Search (2023.emnlp-main)

Copied to clipboard

Haochen Li, Xin Zhou, Anh Luu, Chunyan Miao

CausalDialogue: Modeling Utterance-level Causality in Conversations (2023.findings-acl)

Challenge:	Comparative learning is a key component in fine-tuning code search models . however, negative samples of InfoNCE may deteriorate its representation learning .
Approach:	They propose a loss function that inserts weight terms into InfoNCE to improve contrastive learning.
Outcome:	The proposed loss function is a special case of Soft-InfoNCE, the authors show . it is more accurate than other loss functions, and it is faster than other models.

Copied to clipboard

Yi-Lin Tuan, Alon Albalak, Wenda Xu, Michael Saxon, Connor Pryor, Lise Getoor, William Yang Wang

Dialogue Systems & Conversational Agents

Challenge:	Despite widespread adoption, neural conversation models have yet to exhibit natural chat capabilities with humans . despite their widespread adoption in society, chatbots have yet not shown natural chat capability .
Approach:	They propose a causality-enhanced method to enhance the impact of causality at the utterance level in training neural conversation models.
Outcome:	The proposed method improves diversity and agility of loss functions and still needs improvement . the proposed method is based on a CausalDialogue dataset .

CREST: A Joint Framework for Rationalization and Counterfactual Text Generation (2023.acl-long)

Copied to clipboard

Marcos Treviso, Alexis Ross, Nuno M. Guerreiro, André Martins

Responsible & Trustworthy NLP

Challenge:	Existing methods for analyzing and training NLP models have not been integrated to combine their complementary advantages.
Approach:	They introduce a framework for selective rationalization and counterfactual text generation that leverages CREST to regularize selective rationales and a loss function that regularizes selective rationals.
Outcome:	The proposed framework generates valid counterfactuals that are more natural than those produced by previous methods and can be used for data augmentation at scale.

Learning to Align: Addressing Character Frequency Distribution Shifts in Handwritten Text Recognition (2025.findings-emnlp)

Copied to clipboard

Panagiotis Kaliosis, John Pavlopoulos

Challenge:	Character sets change over time and character frequency distributions shift across historical periods or regions . character distribution alignment can improve existing models at inference time without requiring retraining .
Approach:	They propose a loss function that incorporates the Wasserstein distance between predicted and target distributions.
Outcome:	The proposed method improves accuracy and robustness under temporal and contextual shifts.

Unsupervised Discrete Representations of American Sign Language (2024.emnlp-main)

Copied to clipboard

Artem Abzaliev, Rada Mihalcea

Representation Learning

Challenge:	Modern NLP models use discrete tokens to represent continuous signals, such as videos, audio, or gestures . modalities that are continuous are difficult to use with discrete models, such a LLM .
Approach:	They propose a method that discretizes sequences of fingerspelling signs into tokens . they also propose 'loss function' to improve interpretability of the tokens.
Outcome:	The proposed method improves the performance of the tokenizer on downstream tasks.

Beyond Contrastive Learning: Synthetic Data Enables List-wise Training with Multiple Levels of Relevance (2025.findings-emnlp)

Copied to clipboard

Reza Esfandiarpoor, George Zerveas, Ruochen Zhang, Macton Mgonzo, Carsten Eickhoff, Stephen Bach

Class Distillation with Mahalanobis Contrast: An Efficient Training Paradigm for Pragmatic Language Understanding Tasks (2025.acl-long)

Challenge:	a new approach to training with binary relevance labels uses synthetic data . contrastive learning with binary correlations leaves out subtle nuances useful for ranking .
Approach:	They propose to use waterstein distance as a loss function for training transformer-based retrievers with graduated relevance labels instead of real documents.
Outcome:	The proposed method outperforms conventional training with InfoNCE by a large margin on MARCO and BEIR benchmarks without using real documents.

Copied to clipboard

Chenlu Wang, Weimin Lyu, Ritwik Banerjee