EMPIRICAL ANALYSIS

An Empirical Study on Topic Preservation in Multi-Document Summarization (2022.aacl-srw)

Challenge:	a fundamental characteristic of natural language definitions is that they are widely abundant, pos-1.
Approach:	They propose a multi-relational model that explicitly leverages definitions' semantic structure to derive word embeddings.
Outcome:	The proposed model can preserve the semantic mapping required for interpretable traversal while imposing constraints on definitions while maintaining the recursive semantic structure.

Copied to clipboard

Mong Yuan Sim, Wei Emma Zhang, Congbo Ma

Text Generation

Challenge:	Multi-document summarization is a process of generating an informative and concise summary from multiple topic-related documents.
Approach:	They perform empirical analysis on two MDS datasets and study topic preservation on generated summaries from 8 MDS models.
Outcome:	The results show that extractive and abstractive summarization methods preserve topic information from source documents.

Theoretical Conditions and Empirical Failure of Bracket Counting on Long Sequences with Linear Recurrent Networks (2023.eacl-srw)

Copied to clipboard

Nadine El-Naggar, Pranava Madhyastha, Tillman Weyde

Challenge:	Existing studies have shown that linear RNNs with unbounded activation functions are difficult to train effectively and do not learn exact counting behaviour.
Approach:	They propose to identify the necessary conditions for a linear single-cell RNN to have the ability to count and to investigate how these conditions relate to the empirical behaviour of trained linear RNN models.
Outcome:	The proposed model is a linear single-cell RNN with an unbounded activation function and a Dyck-1-like balanced bracket language.

Sparse, Dense, and Attentional Representations for Text Retrieval (2021.tacl-1)

Copied to clipboard

Yi Luan, Jacob Eisenstein, Kristina Toutanova, Michael Collins

Length-Aware Multi-Kernel Transformer for Long Document Classification (2024.starsem-1)

Challenge:	Dual encoders perform retrieval by encoding documents and queries into dense low-dimensional vectors, scoring each document by its inner product with the query.
Approach:	They propose a dual-encoder-based neural model that combines the efficiency of dual encoders with expressiveness of more costly attentional architectures.
Outcome:	The proposed model outperforms strong alternatives in large-scale retrieval.

Copied to clipboard

Guangzeng Han, Jack Tsao, Xiaolei Huang

Robustness in NLP

Challenge:	Existing SOTA models segment long texts into equal-length snippets, but they have new challenges of context fragmentation and generalizability due to sentence boundaries and varying text lengths.
Approach:	They propose a Length-Aware Multi-Kernel Transformer to encode long documents by transformers and vectorize text length by the kernels to promote model robustness over varying document lengths.
Outcome:	The proposed model outperforms existing models on five benchmarks from health and law domains up to an absolute 10.9% improvement.

LED: A Dataset for Life Event Extraction from Dialogs (2023.findings-eacl)

Copied to clipboard

Yi-Pei Chen, An-Zi Yen, Hen-Hsen Huang, Hideki Nakayama, Hsin-Hsi Chen

Information Extraction & Text Mining

Challenge:	Existing methods for extracting life events from conversations are limited.
Approach:	They propose a dataset containing fine-grained life event annotations on conversational data.
Outcome:	The proposed dataset combines three information extraction frameworks to extract life events from conversations.

Exploring Model Kinship for Merging Large Language Models (2025.findings-emnlp)

Copied to clipboard

Yedi Hu, Yunzhi Yao, Ningyu Zhang, Huajun Chen, Shumin Deng

PAL: Persona-Augmented Emotional Support Conversation Generation (2023.findings-acl)

Challenge:	Model merging has become one of the key technologies for enhancing the capabilities and efficiency of Large Language Models.
Approach:	They propose a model merging strategy that incorporates model kinship to improve model performance.
Outcome:	The proposed model merging strategy can yield better performance on benchmark datasets.

Copied to clipboard

Jiale Cheng, Sahand Sabour, Hao Sun, Zhuang Chen, Minlie Huang

Learning Source Phrase Representations for Neural Machine Translation (2020.acl-main)

Challenge:	Recent work has demonstrated the effectiveness of dialogue models in providing emotional support due to the lack of human resources for mental health support.
Approach:	They propose a framework for dynamically inferring and modeling seekers’ persona from the conversation history and a model that leverages persona information to provide personalized emotional support.
Outcome:	The proposed model outperforms baseline models on the studied benchmark.

Copied to clipboard

Hongfei Xu, Josef van Genabith, Deyi Xiong, Qiuhui Liu, Jingyi Zhang

HybGRAG: Hybrid Retrieval-Augmented Generation on Textual and Relational Knowledge Bases (2025.acl-long)

Challenge:	Existing approaches to machine translation have been shown to be effective for long sentences . however, the attentional network can't capture long-distance dependencies .
Approach:	They propose a multi-head attention mechanism which generates phrase representations from token representations and incorporates them into the Transformer translation model to enhance its ability to capture long-distance relationships.
Outcome:	The proposed model can be computed in parallel and improves on the WMT 14 tasks.

Copied to clipboard

Meng-Chieh Lee, Qi Zhu, Costas Mavromatis, Zhen Han, Soji Adeshina, Vassilis N. Ioannidis, Huzefa Rangwala, Christos Faloutsos

Knowledge Representation

Challenge:	Existing methods for retrieving information from a semi-structured knowledge base are struggling with hybrid questions.
Approach:	They propose a retrieval method that leverages both textual and relational information from a semi-structured knowledge base to answer user questions.
Outcome:	The proposed method surpasses all baselines on the STaRK benchmark and achieves significant performance gains.

Revisiting the Compositional Generalization Abilities of Neural Sequence Models (2022.acl-short)

Copied to clipboard

Arkil Patel, Satwik Bhattamishra, Phil Blunsom, Navin Goyal

Challenge:	Existing studies have suggested that standard seq-to-seq models lack the ability to generalize compositionally.
Approach:	They propose to use one-shot primitive generalization as introduced by the popular SCAN benchmark to modify the training distribution in simple and intuitive ways to achieve near-perfect generalization performance.
Outcome:	The proposed model achieves near-perfect generalization performance despite a lack of training data .

Reframing Instructional Prompts to GPTk’s Language (2022.findings-acl)

Copied to clipboard

Swaroop Mishra, Daniel Khashabi, Chitta Baral, Yejin Choi, Hannaneh Hajishirzi

SimpleQuestions Nearly Solved: A New Upperbound and Baseline Approach (D18-1)

Challenge:	Using reframing techniques, we find that instructional prompts are easier to follow for Language Models (LMs)
Approach:	They propose reframing techniques for manual reformulation of prompts into more effective ones . they compare performance of LMs prompted with reframed instructions on 12 NLP tasks .
Outcome:	The reframing techniques used for prompt reformulation improve performance on 12 tasks . the techniques boost performance on LMs with different sizes compared with original prompts .

Copied to clipboard

Michael Petrochuk, Luke Zettlemoyer

Challenge:	ambiguity in the data bounds performance of the SimpleQuestions dataset, which is commonly used for factoid questions . ambiguities are a problem because many questions have more than one equally plausible interpretation .
Approach:	They propose a benchmark that can be solved by standard methods using the SimpleQuestions dataset . they propose ambiguity in the data bounds performance at 83.4% and a baseline that sets a new state-of-the-art performance level at 78.1% accuracy .
Outcome:	The SimpleQuestions dataset is one of the most commonly used benchmarks for studying factoids . the new benchmark is 78.1% accurate, and the upperbound is loose, the authors show .

ALERT: Adapt Language Models to Reasoning Tasks (2023.acl-long)

Copied to clipboard

Ping Yu, Tianlu Wang, Olga Golovneva, Badr AlKhamissi, Siddharth Verma, Zhijing Jin, Gargi Ghosh, Mona Diab, Asli Celikyilmaz

CoMAE: A Multi-factor Hierarchical Framework for Empathetic Response Generation (2021.findings-acl)

Challenge:	Large language models have shown increasing in-context learning capabilities with scaling up the model and data sizes.
Approach:	They propose a benchmark and suite of analyses to evaluate reasoning skills of large language models.
Outcome:	The proposed model compares pre-trained and fine-tuned models on tasks that require reasoning skills to solve.

Copied to clipboard

Chujie Zheng, Yong Liu, Wei Chen, Yongcai Leng, Minlie Huang

Semantic Sentence Matching via Interacting Syntax Graphs (2022.coling-1)

Challenge:	Existing methods for empathetic response generation ignore hierarchical relationships between different factors, leading to a weak ability of empathy modeling.
Approach:	They propose a multi-factor hierarchical framework for empathetic response generation which models the above three key factors in a hierarchically structured way.
Outcome:	The proposed model generates more empathetic responses than previous methods.

Copied to clipboard

Chen Xu, Jun Xu, Zhenhua Dong, Ji-Rong Wen

Multimodality

Challenge:	Extensive research efforts have been devoted to the task of matching two natural language sentences.
Approach:	They propose to embed syntactic structures into an embedding vector and combine them with other features to predict matching scores.
Outcome:	The proposed method outperforms the state-of-the-art methods on three public datasets and can interpret sentences in interpretable way.

Improving Deep Transformer with Depth-Scaled Initialization and Merged Attention (D19-1)

Copied to clipboard

Biao Zhang, Ivan Titov, Rico Sennrich

Why Machine Reading Comprehension Models Learn Shortcuts? (2021.findings-acl)

Challenge:	Existing methods to improve NLP convergence and computational overhead are limited by stacking more layers.
Approach:	They propose a depth-scaled initialization method which reduces parameter variance at initialization and reduces output variance of residual connections to ease gradient back-propagation.
Outcome:	The proposed method outperforms the base model on translation tasks with five translation directions while matching the decoding speed of the baseline model.

Copied to clipboard

Yuxuan Lai, Chen Zhang, Yansong Feng, Quzhe Huang, Dongyan Zhao

Machine Reading Comprehension

Challenge:	Existing studies show that many MRC models learn shortcuts to outwit benchmarks, but the performance is unsatisfactory in real-world applications.
Approach:	They propose to use shortcut questions to analyze learning difficulty of MRC models . they propose to analyze the learning difficulty regarding shortcut and challenging questions .
Outcome:	The proposed methods show that a large proportion of shortcut questions in training data make models rely on shortcut tricks excessively.

CHIMERA: A Knowledge Base of Scientific Idea Recombinations for Research Analysis and Ideation (2026.acl-long)

Copied to clipboard

Noy Sternlicht, Tom Hope

Knowledge Representation

Challenge:	a hallmark of human innovation is recombination.
Approach:	They propose a task to extract recombination instances from scientific literature . they analyze patterns of recombined concepts and apply it to a broad corpus of AI papers .
Outcome:	The proposed model can predict cross-disciplinary research directions . it can predict recombinations across areas and link methods and concepts .

AnyMAL: An Efficient and Scalable Any-Modality Augmented Language Model (2024.emnlp-industry)

Copied to clipboard

Seungwhan Moon, Andrea Madotto, Zhaojiang Lin, Tushar Nagarajan, Matt Smith, Shashank Jain, Chun-Fu Yeh, Prakash Murugesan, Peyman Heidari, Yue Liu, Kavya Srinet, Babak Damavandi, Anuj Kumar

Mechanistic Interpretability Should Prioritize Feature Consistency in Sparse Autoencoders (2026.acl-long)

Challenge:	Prior work on LLMs focused on models that combine text and one other modality, such as image encoders or proprietary models that are not open sourced.
Approach:	They propose a unified model that reasons over diverse input modality signals and generates textual responses.
Outcome:	The proposed model performs better on multimodal tasks than industry-leading models .

Copied to clipboard

Xiangchen Song, Aashiq Muhamed, Yujia Zheng, Lingjing Kong, Zeyu Tang, Mona T. Diab, Virginia Smith, Kun Zhang

Explainability & Interpretability in NLP

Challenge:	Sparse Autoencoders (SAEs) are a tool in mechanistic interpretability (MI) but the aspiration to identify a canonical set of features is challenged by the observed inconsistency of learned SAE features across different training runs.
Approach:	They propose to use the Pairwise Dictionary Mean Correlation Coefficient to quantify SAE feature consistency as an evaluation axis alongside reconstruction and sparsity.
Outcome:	The proposed measure is based on the pairwise dictionary mean correlation coefficient (PW-MCC) on LLM activations.

Low-resource Entity Set Expansion: A Comprehensive Study on User-generated Text (2022.findings-naacl)

Copied to clipboard

Yutong Shao, Nikita Bhutani, Sajjadur Rahman, Estevam Hruschka

Low-Resource NLP

Challenge:	Existing benchmarks for entity set expansion (ESE) are limited to well-formed text and well-defined concepts.
Approach:	They propose to use user-generated text to assess the generalizability of ESE methods by identifying phenomena such as non-named entities, multifaceted entities and vague concepts.
Outcome:	The proposed methods are based on user-generated text to assess their generalizability and performance.

Perceptual Structure in the absence of grounding: the impact of abstractedness and subjectivity in color language for LLMs (2023.findings-emnlp)

Copied to clipboard

Pablo Loyola, Edison Marrese-Taylor, Andres Hoyos-Idrobo

Challenge:	Existing studies show that color perception and color language are suitable for empirically studying the problem.
Approach:	They propose to quantify alignment between a defined color space and a feature space in a language model by learning a mapping between embedding space and color space.
Outcome:	The results show that there is considerable alignment between a defined color space and the feature space defined by a language model.

Answering Open-Domain Multi-Answer Questions via a Recall-then-Verify Framework (2022.acl-long)

Copied to clipboard

Zhihong Shao, Minlie Huang

Information Retrieval

Challenge:	Existing approaches to open-domain question answering use a rerank-then-read framework . existing approaches use reranked evidence to predict multiple valid answers .
Approach:	They propose to use a recall-then-verify framework to solve open-domain questions . the framework separates the reasoning process of each answer to make better use of retrieved evidence .
Outcome:	The proposed framework predicts significantly more gold answers on open-domain questions than existing systems that use an oracle reranker.

Intent Identification and Entity Extraction for Healthcare Queries in Indic Languages (2023.findings-eacl)

Copied to clipboard

Ankan Mullick, Ishani Mondal, Sourjyadip Ray, Raghav R, G Chaitanya, Pawan Goyal

Information Extraction & Text Mining

Challenge:	Currently, there is a lack of data and technology for resource-poor languages in developing countries like India.
Approach:	They propose to use two different datasets to analyze query intents and entities in healthcare.
Outcome:	The proposed model is useful to identify query intents and entities in real-world scenarios.

A Two-Stage Prediction-Aware Contrastive Learning Framework for Multi-Intent NLU (2024.lrec-main)

Copied to clipboard

Guanhua Chen, Yutong Yao, Derek F. Wong, Lidia S. Chao

Faithful Low-Resource Data-to-Text Generation through Cycle Training (2023.acl-long)

Challenge:	Multi-intent natural language understanding (NLU) models lack the rich information between the shared intents, especially in low-data scenarios.
Approach:	They propose a two-stage framework for multi-intent natural language understanding to harness shared intent information by word-level pre-training and prediction-aware contrastive fine-tuning.
Outcome:	The proposed framework surpasses baselines on low-data and full-data scenarios.

Copied to clipboard

Zhuoer Wang, Marcus Collins, Nikhita Vedula, Simone Filice, Shervin Malmasi, Oleg Rokhlenko

Text Generation

Challenge:	Methods to generate text from structured data have advanced significantly in recent years, but can fail to produce output faithful to the input data, especially on out-of-domain data.
Approach:	They evaluate the effectiveness of cycle training by using two models which are inverses of each other to generate text from structured data and one which generates the structured data from natural language text.
Outcome:	The proposed approach achieves nearly the same performance as fully supervised approaches on the WebNLG, E2E, WTQ, and WSQL datasets.

Hierarchical Recurrent Aggregative Generation for Few-Shot NLG (2022.findings-acl)

Copied to clipboard

Giulio Zhou, Gerasimos Lampouras, Ignacio Iacobacci

Upsample or Upweight? Balanced Training on Heavily Imbalanced Datasets (2025.naacl-long)

Challenge:	Existing approaches do not account for the fact that some sub-tasks, specifically aggregation and lexicalisation, can benefit from transfer learning in different extents.
Approach:	They propose a hierarchical approach for few-shot and zero-shot generation using a three-moduled jointly trained architecture.
Outcome:	The proposed approach achieves state-of-the-art on few-shot and zero-shot settings compared to previous approaches.

Copied to clipboard

Tianjian Li, Haoran Xu, Weiting Tan, Kenton Murray, Daniel Khashabi

ROSE: Robust Selective Fine-tuning for Pre-trained Language Models (2022.emnlp-main)

Challenge:	a lack of data across domains creates significant imbalances in training data sizes . a recent study shows that temperature sampling and scaling are equivalent but differ under stochastic gradient descent due to differences in gradient variance.
Approach:	They propose a method that upsamples low-resource languages and upweights their loss functions to address this disparity.
Outcome:	The proposed method competes effectively with existing data re-weighting techniques while offering computational efficiency.

Copied to clipboard

Lan Jiang, Hao Zhou, Yankai Lin, Peng Li, Jie Zhou, Rui Jiang

Robustness in NLP

Challenge:	Recent studies have highlighted the lack of adversarial robustness in pre-trained models.
Approach:	They propose a fine-tuning approach that conducts selective updates when adapting pre-trained models to downstream tasks.
Outcome:	The proposed approach improves adversarial robustness on downstream tasks . it eliminates spurious updates, leading to flatter and wider optima than the conventional method .

GPT-RE: In-context Learning for Relation Extraction using Large Language Models (2023.emnlp-main)

Copied to clipboard

Zhen Wan, Fei Cheng, Zhuoyuan Mao, Qianying Liu, Haiyue Song, Jiwei Li, Sadao Kurohashi

Do Syntax Trees Help Pre-trained Transformers Extract Information? (2021.eacl-main)

Challenge:	Existing approaches to in-context learning (ICL) are lacking in relation extraction (RE) . emergence of large language models (LLMs) such as GPT-3 represents a significant advancement in natural language processing.
Approach:	They propose to incorporate task-aware representations into demonstration retrieval and enrich the demonstrations with gold label-induced reasoning logic.
Outcome:	The proposed model achieves SOTA and competitive performances on the Semeval and SciERC datasets.

Copied to clipboard

Devendra Sachan, Yuhao Zhang, Peng Qi, William L. Hamilton

Syntactic Text Processing

Challenge:	Recent work suggests that incorporating syntax information from dependency trees can improve task-specific transformer models.
Approach:	They propose to incorporate dependency tree information into pre-trained transformers for three tasks . they propose a late fusion approach and a joint fusion technique to infuses syntax structure into attention layers.
Outcome:	The proposed models obtain state-of-the-art results on SRL and relation extraction tasks.

Consistent Autoformalization for Constructing Mathematical Libraries (2024.emnlp-main)

Copied to clipboard

Lan Zhang, Xin Quan, Andre Freitas

Syntactic Text Processing

Challenge:	Autoformalization is the task of automatically translating mathematical content written in natural language to a formal language expression.
Approach:	They propose to use three mechanisms to improve autoformalization quality . they propose to combine most-similar retrieval augmented generation, denoising steps and auto-correction with syntax error feedback to improve syntactic, terminological and semantic control.
Outcome:	The proposed mechanisms can deliver syntactically, terminologically and semantically more consistent results across different models.

Re3: A Holistic Framework and Dataset for Modeling Collaborative Document Revision (2024.acl-long)

Copied to clipboard

Qian Ruan, Ilia Kuznetsov, Iryna Gurevych

Challenge:	a framework for collaborative document revision is lacking for empirical analysis and NLP.
Approach:	They propose a framework for joint analysis of collaborative document revision that instantiates a corpus of aligned scientific paper revisions manually labeled according to their action and intent.
Outcome:	The proposed framework provides first empirical insights into collaborative document revision in the academic domain and assesses its capabilities.

Select-Then-Decompose: From Empirical Analysis to Adaptive Selection Strategy for Task Decomposition in Large Language Models (2025.emnlp-main)

Copied to clipboard

Shuodi Liu, Yingzhuo Liu, Zi Wang, Yusheng Wang, Huijia Wu, Liuyu Xiang, Zhaofeng He

Visually-Guided Policy Optimization for Multimodal Reasoning (2026.acl-long)

Challenge:	Existing task decomposition methods focus on memory, tool usage, and feedback mechanisms, but they often overlook the trade-off between performance and cost.
Approach:	They propose a strategy that selects the most suitable decomposition approach based on task characteristics and enhances the reliability of the results through a verification module.
Outcome:	The proposed strategy is based on categories of approaches, characteristics of tasks, and configuration of decomposition and execution models.

Copied to clipboard

Zengbin Wang, Feng Xiong, Liang Lin, Xuecai Hu, Yong Wang, Yanlin Wang, Man Zhang, Xiangxiang Chu

Multimodality

Challenge:	Existing RLVRs lack visual faithfulness due to text-dominated reasoning . a novel framework to reinforce visual focus during policy optimization is proposed .
Approach:	They propose a framework to reinforce visual focus during policy optimization using visual attention compensation mechanism.
Outcome:	The proposed framework exhibits better visual activation and superior performance in multimodal reasoning and visual-dependent tasks.

MSD: Saliency-aware Knowledge Distillation for Multimodal Understanding (2021.findings-emnlp)

Copied to clipboard

Woojeong Jin, Maziar Sanjabi, Shaoliang Nie, Liang Tan, Xiang Ren, Hamed Firooz

Is ChatGPT a Good Multi-Party Conversation Solver? (2023.findings-emnlp)

Challenge:	Current knowledge distillation models are limited and lack performance on multimodal datasets.
Approach:	They propose a multimodal knowledge distillation framework to transfer knowledge from a teacher on multimodal tasks by learning the teacher's behavior within each modality.
Outcome:	The proposed framework achieves better performance than KD on four multimodal datasets.

Copied to clipboard

Chao-Hong Tan, Jia-Chen Gu, Zhen-Hua Ling

Unveiling Selection Biases: Exploring Order and Token Sensitivity in Large Language Models (2024.findings-acl)

Challenge:	Large Language Models (LLMs) are powerful tools for multi-party conversations, but their capacity to handle multi-parties remains unexplored.
Approach:	They propose to evaluate ChatGPT and GPT-4's zero-shot learning capabilities within the context of multi-party conversations (MPCs) they also propose to incorporate MPC structures, encompassing both speaker and addressee architecture.
Outcome:	The proposed models perform poorly on a number of MPC tasks while GPT-4 performs well on speaker and addressee architecture.

Copied to clipboard

Sheng-Lun Wei, Cheng-Kuang Wu, Hen-Hsen Huang, Hsin-Hsi Chen

Unlocking General Long Chain-of-Thought Reasoning Capabilities of Large Language Models via Representation Engineering (2025.acl-long)

Challenge:	Using zero-shot or few-shot prompting, Large Language Models have been widely adopted in downstream applications.
Approach:	They propose to quantify the impact of option order and token usage on LLMs and propose mitigation strategies to enhance model performance.
Outcome:	The proposed mitigation strategies improve model performance and reduce the impact of token and order sensitivity on LLMs.

Copied to clipboard

Xinyu Tang, Xiaolei Wang, Zhihao Lv, Yingqian Min, Xin Zhao, Binbin Hu, Ziqi Liu, Zhiqiang Zhang

Frustratingly Easy Label Projection for Cross-lingual Transfer (2023.findings-acl)

Challenge:	Existing work finds that long CoT reasoning can be efficiently elicited by tuning on only a few examples and can easily transfer to other tasks.
Approach:	They propose a representation engineering method to unleash the general long CoT reasoning capabilities of LLMs.
Outcome:	The proposed method is effective in in-domain and cross-domain scenarios.

Copied to clipboard

Yang Chen, Chao Jiang, Alan Ritter, Wei Xu

Multilinguality

Challenge:	Existing approaches to improve cross-lingual transfer performance are based on word alignment, but no empirical studies have evaluated their effectiveness or limitations.
Approach:	They propose a mark-then-translate method that integrates translation and projection by inserting special markers around the labeled spans in the original sentence.
Outcome:	The proposed method outperforms word alignment-based methods in 57 languages and three tasks.

Self-Supervised Behavior Cloned Transformers are Path Crawlers for Text Games (2023.findings-emnlp)

Copied to clipboard

Ruoyao Wang, Peter Jansen

How Effective is Task-Agnostic Data Augmentation for Pretrained Transformers? (2020.findings-emnlp)

Challenge:	Text game agents are often modeled using reinforcement learning, but their performance is limited.
Approach:	They propose a self-supervised behavior cloning transformer for text games . they explore trajectories that lead to reward within the games and then train small models . their approach consistently uncovers generalizable training data, achieving 90% performance of supervised systems across three benchmark text games.
Outcome:	The proposed model achieves 90% performance on three text games.

Copied to clipboard

Shayne Longpre, Yu Wang, Chris DuBois

DLPO: Towards a Robust, Efficient, and Generalizable Prompt Optimization Framework from a Deep-Learning Perspective (2025.findings-emnlp)

Challenge:	Task-agnostic data augmentations have proven widely effective in computer vision, even on pretrained models.
Approach:	They examine the effects of two types of task-agnostic data augmentation on pretrained transformers using 5 classification tasks and 6 datasets.
Outcome:	The proposed techniques improve performance on 5 classification tasks, 6 datasets, and 3 variants of modern pretrained transformers.

Copied to clipboard

Dengyun Peng, Yuhang Zhou, Qiguang Chen, JinHao Liu, Jingjing Chen, Libo Qin, Wanxiang Che

DialogSum: A Real-Life Scenario Dialogue Summarization Dataset (2021.findings-acl)

Challenge:	Existing methods for prompt optimization still face challenges in robustness, efficiency, and generalization.
Approach:	They propose 7 new approaches inspired by traditional deep learning paradigms for prompt optimization that integrate text-based gradient optimization.
Outcome:	The proposed methods integrate deep learning paradigms into text-based gradient optimization.

Copied to clipboard

Yulong Chen, Yang Liu, Liang Chen, Yue Zhang

Efficient Nearest Neighbor Language Models (2021.emnlp-main)

Challenge:	Experimental results show unique challenges in dialogue summarization such as spoken terms, special discourse structures, coreferences and ellipsis, pragmatics and social common sense.
Approach:	They propose a large-scale labeled dialogue summarization dataset . they use state-of-the-art neural models to analyze spoken dialogue summaries .
Outcome:	The proposed dataset can be used to analyze spoken dialogue summarization challenges.

Copied to clipboard

Junxian He, Graham Neubig, Taylor Berg-Kirkpatrick

CMIE: Combining MLLM Insights with External Evidence for Explainable Out-of-Context Misinformation Detection (2025.findings-acl)

Challenge:	Non-parametric neural language models (NLMs) learn text distributions by memorizing training data points.
Approach:	They propose to use an external datastore to learn from a non-parametric language model.
Outcome:	The proposed methods achieve up to a 6x speed-up in inference speed while retaining comparable performance.

Copied to clipboard

Fanxiao Li, Jiaying Wu, Canyuan He, Wei Zhou

Boundary Smoothing for Named Entity Recognition (2022.acl-long)

Challenge:	Multimodal large language models have demonstrated impressive capabilities in visual reasoning and text generation.
Approach:	They propose a multimodal large language model that captures deeper relationships between images and text . they propose CMIE, which uses a Coexistence Relationship Generation strategy and an AS mechanism to detect misinformation.
Outcome:	The proposed framework outperforms existing methods in detecting out-of-context misinformation.

Copied to clipboard

Enwei Zhu, Jinpeng Li

Information Extraction & Text Mining

Challenge:	Named entity recognition models often encounter over-confidence issues . boundary smoothing is a method that re-assigns entity probabilities from annotated spans to the surrounding ones .
Approach:	They propose a method for regularizing entity probabilities from annotated spans to the surrounding ones.
Outcome:	The proposed method achieves better than or competitive with previous state-of-the-art systems on well-known benchmarks.

Synthetic Pre-Training Tasks for Neural Machine Translation (2023.findings-acl)

Copied to clipboard

Zexue He, Graeme Blackwood, Rameswar Panda, Julian McAuley, Rogerio Feris

Multilinguality

Challenge:	toxicity and bias can be addressed by pre-training with synthetic resources . BLEU scores are used to compare methods with real-world data .
Approach:	They propose several ways to generate obfuscated data from large parallel corpus and concatenating phrase pairs from small word-aligned corpus with synthetic parallel data without real human language corpora.
Outcome:	The proposed methods can be used to generate obfuscated data or synthetic parallel data without real human language corpora even with high levels of oblication.

Towards Persona-Based Empathetic Conversational Models (2020.emnlp-main)

Copied to clipboard

Peixiang Zhong, Chen Zhang, Hao Wang, Yong Liu, Chunyan Miao

Chain-of-Memory: Lightweight Memory Construction with Dynamic Evolution for LLM Agents (2026.acl-long)

Challenge:	Empathetic conversational models have been shown to improve user satisfaction and task outcomes in numerous domains.
Approach:	They propose a task towards persona-based empathetic conversations and propose e-learning model CoBERT that can be used to train persona on emmpathetic conversations.
Outcome:	The proposed model improves empathetic responding more when trained on e-mpathetic conversations than non-empathy ones.

Copied to clipboard

Xiucheng Xu, Bingbing Xu, Tian Xueyun, Zihe Huang, Rongxin Chen, Li Yunfan, Huawei Shen

Information Retrieval

Challenge:	Existing paradigms for large language model (LLM) agents use memory construction and retrieval-augmented generation.
Approach:	They propose a framework that advocates for a paradigm shift toward lightweight construction paired with sophisticated utilization.
Outcome:	Experiments show that CoM outperforms baselines with accuracy gains of 7.5%–10.4% while reducing computational overhead to approximately 2.7% of token consumption and 6.0% of latency compared to complex memory architectures.

DISK: Domain-constrained Instance Sketch for Math Word Problem Generation (2022.coling-1)

Copied to clipboard

Tianyang Cao, Shuang Zeng, Xiaodan Xu, Mairgup Mansur, Baobao Chang

Zero-Shot Classification by Logical Reasoning on Natural Language Explanations (2023.findings-acl)

Challenge:	Existing methods for generating MWP text from equations are inflexible and require pre-defined templates.
Approach:	They propose a neural model which generates MWPs from equations by constructing a Quantity Cell Graph from the retrieved MWp instance and reasoning over it.
Outcome:	The proposed model performs impressively on educational MWP set and on human evaluation metrics.

Copied to clipboard

Chi Han, Hengzhi Pei, Xinya Du, Heng Ji

Instance-Based Learning of Span Representations: A Case Study through Named Entity Recognition (2020.acl-main)

Challenge:	Experimental results show that CLORE is superior to baselines on zero-shot classification tasks.
Approach:	They propose a framework for classification by logically parsing and reasoning on natural language explanations.
Outcome:	The proposed framework outperforms baselines on zero-shot classification tasks.

Copied to clipboard

Hiroki Ouchi, Jun Suzuki, Sosuke Kobayashi, Sho Yokoi, Tatsuki Kuribayashi, Ryuto Konno, Kentaro Inui

CRaSh: Clustering, Removing, and Sharing Enhance Fine-tuning without Full Large Language Model (2023.emnlp-main)

Challenge:	Recent neural networks can induce good span feature representations and achieve high performance in structured prediction tasks.
Approach:	They propose an instance-based learning method that learns similarities between spans . they aim to build models that have high interpretability without sacrificing performance .
Outcome:	The proposed method improves interpretability without sacrificing performance.

Copied to clipboard

Kaiyan Zhang, Ning Ding, Biqing Qi, Xuekai Zhu, Xinwei Long, Bowen Zhou

Rethinking Masked Language Modeling for Chinese Spelling Correction (2023.acl-long)

Challenge:	Instruction tuning is an effective way of aligning large language models with private instruction data.
Approach:	They propose a training-free strategy to derive improved emulators from LLMs by using Offsite-Tuning (OFT) they propose CRaSh, which transfers transformer blocks between centralized LLM and downstream emulators .
Outcome:	The proposed technique boosts performance of large language models with billions of parameters.

Copied to clipboard

Hongqiu Wu, Shaohua Zhang, Yuchen Zhang, Hai Zhao

What is lost in Normalization? Exploring Pitfalls in Multilingual ASR Model Evaluations (2024.emnlp-main)

Challenge:	Existing CSC models over-fit the error model while under-fitting the language model, resulting in poor generalization to out-of-distribution error patterns.
Approach:	They propose to use a multi-domain benchmark LEMON to assess the open domain generalization of Chinese Spelling Correction models.
Outcome:	The proposed method achieves state-of-the-art results on SIGHAN, ECSpell, and LEMON.

Copied to clipboard

Kavya Manohar, Leena Pillai

Speech & Audio in NLP

Challenge:	Existing text normalization routines that target Indic scripts are flawed when applied to multilingual automatic speech recognition models.
Approach:	They propose to develop text normalization routines that leverage native linguistic expertise to ensure more robust and accurate evaluations of multilingual automatic speech recognition models.
Outcome:	The proposed normalization routines can be leveraged to improve performance metrics for Indic languages.

Semantic Role Labeling as Syntactic Dependency Parsing (2020.emnlp-main)

Copied to clipboard

Tianze Shi, Igor Malioutov, Ozan Irsoy

Syntactic Text Processing

Challenge:	Using propBank-style semantic role labeling, we reduce the task to syntactic dependency parsing.
Approach:	They propose to convert SRL annotations into dependency tree representations through joint labels that permit highly accurate recovery back to the original format.
Outcome:	The proposed scheme reduces the task of (span-based) PropBank-style semantic role labeling to syntactic dependency parsing.

Fine-tuning Smaller Language Models for Question Answering over Financial Documents (2024.findings-emnlp)

Copied to clipboard

Karmvir Phogat, Sai Akhil Puranam, Sridhar Dasaratha, Chetan Harsha, Shashishekar Ramakrishna

Question Answering

Challenge:	Recent research has shown that smaller language models can acquire substantial reasoning abilities when fine-tuned with reasoning exemplars crafted by a significantly larger teacher model.
Approach:	They propose to fine-tune several smaller model to generate programs that encode the required financial reasoning and calculations.
Outcome:	The proposed model outperforms the teacher model in the financial domain by adjusting the entity extraction for the specific data format.

Making Pre-trained Language Models both Task-solvers and Self-calibrators (2023.findings-acl)

Copied to clipboard

Yangyi Chen, Xingyao Wang, Heng Ji

Slamming: Training a Speech Language Model on One GPU in a Day (2025.findings-acl)

Challenge:	Existing work shows that pre-trained language models can be effective for high-stake applications, but they become overconfident in their wrong predictions.
Approach:	They propose to use extra data to train pre-trained language models to effectively utilize training samples to make them both task-solvers and self-calibrators.
Outcome:	The proposed method can be used in three downstream applications, including selective classification, adversarial defense, and model cascading.

Copied to clipboard

Gallil Maimon, Avishai Elmakies, Yossi Adi

NEWTON: Are Large Language Models Capable of Physical Reasoning? (2023.findings-emnlp)

Challenge:	Slam is a recipe for training high-quality Speech Language Models (SLMs) on a single academic GPU in 24 hours.
Approach:	They propose a recipe for training high-quality Speech Language Models on a single academic GPU in 24 hours.
Outcome:	The proposed training recipe outperforms predicted compute optimal performance, giving an optimistic view to SLM feasibility.

Copied to clipboard

Yi Wang, Jiafei Duan, Dieter Fox, Siddhartha Srinivasa

Multimodal Transformer for Unaligned Multimodal Language Sequences (P19-1)

Challenge:	Large Language Models have been shown to encapsulate syntactic, semantic, word sense, and common-sense knowledge, but limited exploration of their physical reasoning abilities has been conducted.
Approach:	They propose a repository and benchmark to evaluate LLMs' physical reasoning skills . they use a pipeline to generate a variant of the benchmark customized to the objects and attributes relevant for their application.
Outcome:	The proposed benchmark examines the reasoning capabilities of language models across reasoning tasks.

Copied to clipboard

Yao-Hung Hubert Tsai, Shaojie Bai, Paul Pu Liang, J. Zico Kolter, Louis-Philippe Morency, Ruslan Salakhutdinov

Efficient Entity Candidate Generation for Low-Resource Languages (2022.lrec-1)

Challenge:	Human language is often multimodal, which comprehends a mixture of natural language, facial gestures, and acoustic behaviors.
Approach:	They propose a multimodal model that extends the standard Transformer network to learn representations directly from unaligned multimodal streams.
Outcome:	The proposed model outperforms state-of-the-art methods on aligned and non-aligned data.

Copied to clipboard

Alberto Garcia-Duran, Akhil Arora, Robert West

SLANG: New Concept Comprehension of Large Language Models (2024.emnlp-main)

Challenge:	Existing approaches for cross-lingual entity linking are not suitable for English.
Approach:	They propose a candidate generation problem in cross-lingual entity linking with a focus on low-resource languages.
Outcome:	The proposed solution outperforms the state-of-the-art approach on 9 real-world datasets and query types.

Copied to clipboard

Lingrui Mei, Shenghua Liu, Yiwei Wang, Baolong Bi, Xueqi Cheng

ProFit: Leveraging High-Value Signals in SFT via Probability-Guided Token Selection (2026.findings-acl)

Challenge:	Dynamic nature of language limits the adaptability of Large Language Models (LLMs) Traditionally, LLMs are trained on static data, which limits their adaptability .
Approach:	They propose a benchmark to integrate novel data and assess LLMs’ ability to comprehend emerging concepts, alongside a causal inference-based approach to enhance LLM comprehension of new phrases and their colloquial context.
Outcome:	The proposed model outperforms baseline models in terms of precision and relevance in the comprehension of Internet slang and memes.

Copied to clipboard

Tao Liu, Taiqiang Wu, Runming Yang, Shaoning Sun, Junjie Wang, Yujiu Yang

Challenge:	Traditional fine-tuning ignores one-to-many nature of language, leading to overfitting . authors propose a method to fine- tune LLMs by leveraging tokens.
Approach:	They propose a method to fine-tune Large Language Models by leveraging tokens to mask low-probability tokens.
Outcome:	The proposed method outperforms baselines on general reasoning and mathematical benchmarks.

SelfAug: Mitigating Catastrophic Forgetting in Retrieval-Augmented Generation via Distribution Self-Alignment (2025.findings-emnlp)

Copied to clipboard

Yuqing Huang, Rongyang Zhang, Qimeng Wang, Chengqiang Lu, Yan Gao, null Yiwu, Yao Hu, Xuyang Zhi, Guiquan Liu, Xin Li, Hao Wang, Enhong Chen

TUBA: Cross-Lingual Transferability of Backdoor Attacks in LLMs with Instruction Tuning (2025.findings-acl)

Challenge:	Existing solutions for supervised fine-tuning often lead to catastrophic forgetting, where models lose their previously acquired knowledge and general capabilities.
Approach:	They propose a self-distribution alignment method that aligns input sequence logits to preserve the model’s semantic distribution, thereby mitigating catastrophic forgetting and improving downstream performance.
Outcome:	The proposed method achieves a superior balance between downstream learning and general capability retention.

Copied to clipboard

Xuanli He, Jun Wang, Qiongkai Xu, Pasquale Minervini, Pontus Stenetorp, Benjamin I. P. Rubinstein, Trevor Cohn

Two-Stage Fine-Tuning for Improved Bias and Variance for Large Pretrained Language Models (2023.acl-long)

Challenge:	Despite the increasing support for multilingual capabilities, the impact of backdoor attacks on LLMs remains under-explored.
Approach:	They propose to use poisoned instructiontuning data to attack multilingual LLMs . their results show that more powerful models show increased susceptibility to transferable cross-lingual backdoor attacks .
Outcome:	The proposed attack is effective in models like BLOOM and GPT-4o with high success rates in more than 7 out of 12 languages.

Copied to clipboard

Lijing Wang, Yingya Li, Timothy Miller, Steven Bethard, Guergana Savova

Analyzing Transformers in Embedding Space (2023.acl-long)

Challenge:	Recent work challenges the bias-variance trade-off . large pretrained models can have large variance and overfit domain-specific data .
Approach:	They propose a bias-variance trade-off that implies learning methods need to balance complexity with data size to minimize under-fitting and over-fit.
Outcome:	The proposed method achieves strong results on SuperGLUE and clinical information extraction tasks.

Copied to clipboard

Guy Dar, Mor Geva, Ankit Gupta, Jonathan Berant

Data Factors for Better Compositional Generalization (2023.emnlp-main)

Challenge:	Understanding Transformer-based models has attracted significant attention . a zero-pass approach is feasible for some parameters, and for two-layer attention networks .
Approach:	They propose a theoretical framework where parameters of a trained Transformer are interpreted by projecting them into the embedding space.
Outcome:	The proposed framework shows that pre-trained and fine-tuned models can be interpreted in embedding space.

Copied to clipboard

Xiang Zhou, Yichen Jiang, Mohit Bansal

Challenge:	Recent diagnostic datasets on compositional generalization expose severe problems . state-of-the-art models trained on larger and more general datasets show better generalization ability .
Approach:	They conduct an empirical analysis by training Transformer models on a variety of training sets with different data factors including dataset scale, pattern complexity, example difficulty, etc.
Outcome:	The proposed model training on larger datasets improves on compositional generalization tasks.

A Comprehensive Evaluation of Tool-Assisted Generation Strategies (2023.findings-emnlp)

Copied to clipboard

Alon Jacovi, Avi Caciularu, Jonathan Herzig, Roee Aharoni, Bernd Bohnet, Mor Geva

Logic Rules as Explanations for Legal Case Retrieval (2024.lrec-main)

Challenge:	Various few-shot tool-usage strategies have been proposed to overcome LMs' shortcomings.
Approach:	They propose to augment language models with tools to overcome their shortcomings . they find strong no-tool baselines are competitive to tool-assisted strategies .
Outcome:	The proposed strategies outperform those that refine incorrect outputs with tools in knowledge-retrieval tasks, the study finds . the findings suggest few-shot tool integration is still an open challenge .

Copied to clipboard

ZhongXiang Sun, Kepu Zhang, Weijie Yu, Haoyu Wang, Jun Xu

Explainability & Interpretability in NLP

Challenge:	Recent efforts to learn explainable legal case retrieval models fail to provide faithful and interpretable explanations for legal cases.
Approach:	They propose a framework that uses logic rules to explain legal case retrieval results . they extend benchmarks of LeCaRD and ELAM with manually annotated logic rules .
Outcome:	The proposed framework is able to provide faithful explanations for legal case retrieval.

Path Drift in Large Reasoning Models: How First-Person Commitments Override Safety (2025.emnlp-main)

Copied to clipboard

Yuyi Huang, Runzhe Zhan, Lidia S. Chao, Ailin Tao, Derek F. Wong

Can LLMs Help Uncover Insights about LLMs? A Large-Scale, Evolving Literature Analysis of Frontier LLMs (2025.acl-long)

Challenge:	Existing studies on prompt injection and jailbreak attacks primarily target the surface structure of input prompts.
Approach:	They propose a three-stage approach to mitigate the risk of Long-CoT reasoning drift . they propose 'path-level defense' strategy that incorporates role attribution correction and metacognitive reflection .
Outcome:	The proposed framework reduces refusal rates and ethical evaporation, while ethical escalation and layered disclaimers progressively steer models toward unsafe completions.

Copied to clipboard

Jungsoo Park, Junmo Kang, Gabriel Stanovsky, Alan Ritter

Aligning Predictive Uncertainty with Clarification Questions in Grounded Dialog (2023.findings-emnlp)

Challenge:	Recent surveys of literature highlight the overwhelming growth of Large Language Models (LLMs).
Approach:	They propose a semi-automated literature analysis approach that automates literature analysis using LLMs.
Outcome:	The proposed approach reduces paper surveying and data extraction by 93% compared to manual methods.

Copied to clipboard

Kata Naszadi, Putra Manggala, Christof Monz

Natural Language Interfaces

Challenge:	Previous work bases the timing of questions on supervised models learned from interactions between humans.
Approach:	They propose to ground the need for questions in the acting agent's predictive uncertainty by using the T5 encoder-decoder architecture to solve a Minecraft Collaborative Building task.
Outcome:	The proposed model can detect ambiguous instructions and predict responses better than previous models.

Measuring Pointwise 𝒱-Usable Information In-Context-ly (2023.findings-emnlp)

Copied to clipboard

Sheng Lu, Shan Chen, Yingya Li, Danielle Bitterman, Guergana Savova, Iryna Gurevych

Challenge:	In-context learning (ICL) is a new learning paradigm that has gained popularity along with the development of large language models.
Approach:	They propose to adapt a recently proposed hardness metric, pointwise V-usable information (PVI), to an in-context version.
Outcome:	The proposed hardness metric is compared with the original model and is more efficient because it requires only a few exemplars and does not require fine-tuning.

D-QRELO: Training- and Data-Free Delta Compression for Large Language Models via Quantization and Residual Low-Rank Approximation (2026.findings-acl)

Copied to clipboard

Junlin Li, Shuangyong Song, Guodong DU, Ngai Wong, Xuebo Liu, Yongxiang Li, Min Zhang, Jing Li, Xuelong Li

FoE: Forest of Errors Makes the First Solution the Best in Large Reasoning Models (2026.acl-long)

Challenge:	Existing methods for fine-tuned large language models fail on fine-scale datasets . large data scale amplifies delta parameter magnitude, singular values, and entropy, causing compression errors.
Approach:	They propose a training- and data-free delta compression method that captures dominant delta structure and compensates residual low-rank approximation to recover fine-grained details from smaller residual error.
Outcome:	The proposed method outperforms existing methods on large-scale datasets on dense and MoE architectures.

Copied to clipboard

Kehan Jiang, Haonan Dong, Zhaolu Kang, Zhengzhou Zhu, Guojie Song

Memorize Step by Step: Efficient Long-Context Prefilling with Incremental Memory and Decremental Chunk (2024.emnlp-main)

Challenge:	Recent Large Reasoning Models (LRMs) have demonstrated remarkable success in complex reasoning tasks.
Approach:	They propose a self-guided efficient reasoning framework that reduces FoE by pruning subs.
Outcome:	The proposed model outperforms eight competitive baselines while reducing token consumption by 37.7% 70.4%.

Copied to clipboard

Zhiyuan Zeng, Qipeng Guo, Xiaoran Liu, Zhangyue Yin, Wentao Shu, Mianqiu Huang, Bo Wang, Yunhua Zhou, Linlin Li, Qun Liu, Xipeng Qiu

Green & Sustainable NLP

Challenge:	Existing methods to optimize LLM for long sequences for long documents are slow and consume memory.
Approach:	They propose a method that starts with a small memory size and gradually increases it . they propose Decremental Chunk based on Incremental Memory (IMDC) which reduces chunk size while increasing memory size .
Outcome:	The proposed method is faster (1.45x) and reduces GPU memory consumption by 23.3% compared to fixed-size memory.

REPRO-Bench: Can Agentic AI Systems Assess the Reproducibility of Social Science Research? (2025.findings-acl)

Copied to clipboard

Chuxuan Hu, Liyun Zhang, Yeji Lim, Aum Wadhwani, Austin Peters, Daniel Kang

Text Takes Over: A Study of Modality Bias in Multimodal Intent Detection (2025.emnlp-main)

Challenge:	Existing benchmarks for reproducing social science papers focus on reproducing results using provided code and data without assessing their consistency with the paper.
Approach:	They propose a benchmark to evaluate agentic AI systems' ability to automate reproducibility assessment.
Outcome:	The proposed benchmark oversimplifies real-world scenarios and lacks diversity in data formats and programming languages.

Copied to clipboard

Ankan Mullick, Saransh Sharma, Abhik Jana, Pawan Goyal

Sentiment Analysis

Challenge:	a new study examines the effectiveness of large language models and non-LLMs in multimodal intent detection . large-scale multimodal data integrations include text, audio, and visual inputs .
Approach:	They propose a framework to debias multimodal intent detection datasets by using human evaluation.
Outcome:	The proposed framework debiases the datasets and shows that mistral-7B outperforms most competitive models by approximately 9% on MIntRec-1 and 4% on MIndRec2.0.

Training-free Deep Concept Injection Enables Language Models for Video Question Answering (2024.emnlp-main)

Copied to clipboard

Xudong Lin, Manling Li, Richard Zemel, Heng Ji, Shih-Fu Chang

Where Reasoning Breaks: Logic-Aware Path Selection by Controlling Logical Connectives in LLMs Reasoning Chains (2026.findings-acl)

Challenge:	Existing methods to train pretrained language models for zero-shot crossmodal tasks require crossmodal pretraining.
Approach:	They propose to inject visual concepts into the input text embedding space of a pretrained language model and build adaptation layers based on the intermediate representation of concepts.
Outcome:	The proposed model performs zero-shot crossmodal tasks without crossmodal pretraining . it is based on the injection of visual concepts as input tokens and augmentation in intermediate features . the proposed model achieves competitive or even better results in zero- shot and fine-tuning settings .

Copied to clipboard

Seunghyun Park, Yuanyuan Lei

Label Set Optimization via Activation Distribution Kurtosis for Zero-Shot Classification with Generative Models (2025.emnlp-main)

Challenge:	a single transition error can propagate through the entire reasoning chain, leading to unstable performance.
Approach:	They propose a framework that intervenes at logical connective junctions to improve LLMs' reasoning.
Outcome:	The proposed framework achieves favorable accuracy–efficiency trade-off compared to global inference time scaling methods like beam search and self-consistency.

Copied to clipboard

Yue Li, Zhixue Zhao, Carolina Scarton

Probability Distribution Collapse: A Critical Bottleneck to Compact Unsupervised Neural Grammar Induction (2025.emnlp-main)

Challenge:	In-context learning (ICL) performance is highly sensitive to prompt design, yet the impact of class label options (e.g. lexicon or order) in zero-shot classification remains underexplored.
Approach:	They propose a post-hoc method for selecting optimal label sets in zero-shot ICL with large language models.
Outcome:	The proposed method consistently achieves performance gains of 0.54 to 0.76 compared to the conventional method.

Copied to clipboard

Jinwook Park, Kangil Kim

Low-Resource NLP

Challenge:	Existing models face expressiveness bottlenecks, resulting in unnecessarily large yet underperforming grammars.
Approach:	They propose a method to reduce the expressiveness bottleneck of unsupervised neural grammar induction by leveraging neural parameterization to estimate prob-ability distributions.
Outcome:	The proposed approach significantly improves parsing performance while enabling the use of significantly more compact grammars across a wide range of languages.

CCD: Mitigating Hallucinations in Radiology MLLMs via Clinical Contrastive Decoding (2026.findings-acl)

Copied to clipboard

Xi Zhang, Zaiqiao Meng, Jake Lever, Edmond S. L. Ho

Think How to Think: Mitigating Overthinking with Autonomous Difficulty Cognition in Large Reasoning Models (2026.acl-long)

Challenge:	Multimodal large language models generate medical hallucinations due to over-sensitivity to clinical sections.
Approach:	They propose a framework that integrates structured clinical signals from task-specific radiology expert models.
Outcome:	The proposed framework improves overall performance on radiology report generation (RRG) on the MIMIC-CXR dataset, it yields up to 17% improvement in RadGraph-F1.

Copied to clipboard

Yongjiang Liu, Haoxi Li, Xiaosong Ma, Jie Zhang, Song Guo

AI Agents for the Science of Science: A Survey of Tasks, Architectures, Evaluations, and Challenges (2026.findings-acl)

Challenge:	Recent Large Reasoning Models (LRMs) excel at complex reasoning tasks but often suffer from overthinking.
Approach:	They propose a two-stage fine-tuning strategy that progressively inspires LRMs’ difficulty cognition and redundancy cognition of LRM.
Outcome:	The proposed model significantly reduces inference costs by over 70% on easy tasks and 40% on complex ones without compromising performance.

Copied to clipboard

Yixuan Liu, Yicheng Zhang

Challenge:	The Science of Science (SciSc) examines how scientific knowledge is produced, evaluated, and transformed by utilizing large-scale scholarly and bibliometric data.
Approach:	They propose a task-centered taxonomy for AI agents that model citations, collaborations, and community dynamics.
Outcome:	The proposed taxonomy distinguishes agents as simulations from tools for empirical analysis and scientific workflows.

Libra-VLA: Achieving Learning Equilibrium via Asynchronous Coarse-to-Fine Dual-System (2026.acl-long)

Copied to clipboard

Yifei Wei, Linqing Zhong, Yi Liu, Yuxiang Lu, Xindong He, Maoqing Yao, Guanghui Ren

Challenge:	Vision-Language-Action models ground high-level semantic instructions into executable physical actions.
Approach:	They propose a Coarse-to-Fine Dual-System VLA architecture that decouples learning complexity into a coarse-to fine hierarchy while leveraging structural modularity to implement an asynchronous execution strategy.
Outcome:	The proposed architecture decouples learning complexity into a coarse-to-fine hierarchy while leveraging structural modularity to implement an asynchronous execution strategy.

The African Languages Lab: A Collaborative Approach to Advancing Low-Resource African NLP (2026.acl-long)

Copied to clipboard

Sheriff Issaka, Keyi Wang, Yinka Ajibola, Oluwatumininu Samuel-Ipaye, Zhaoyi Zhang, Nicte Aguillon Jimenez, Evans Kofi Agyei, Abraham Lin, Rohan Ramachandran, Sadick Abdul Mumin, Faith Nchifor, Mohammed Shuraim Issah, Erick Rosas Gonzalez, Lieqi Liu, Sylvester Kpei, Jemimah Kusi Osei, Carlene Ajeneza, Persis Boateng, Prisca Adwoa Dufie Yeboah, Saadia Gabriel

Multimodality

Challenge:	Among the approximately 7,000 languages spoken globally, fewer than 20 receive substantial attention in NLP research.
Approach:	They propose to use African multi-modal speech and text data to validate African multimodal models and validate them on targeted language data.
Outcome:	The African Languages Lab's results show that the proposed model outperforms untrained models in 31 languages and a 1B-parameter model beats the commercial system in Yoruba and Twi.

Attention Sinks in Diffusion Language Models (2026.findings-acl)

Copied to clipboard

Maximo Eduardo Rulli, Simone Petruzzi, Edoardo Michielon, Fabrizio Silvestri, Simone Scardapane, Alessio Devoto