Role-playing is important for Large Language Models (LLMs) to follow diverse instructions while maintaining role identity and the role's pre-defined ability limits. Existing role-playing datasets mostly contribute to controlling role style and knowledge boundaries, but overlook role-playing in instruction-following scenarios. We introduce a fine-grained role-playing and instruction-following composite benchmark, named RoleMRC, including: (1) Multi-turn dialogues between ideal roles and humans, including free chats or discussions upon given passages; (2) Role-playing machine reading comprehension, involving response, refusal, and attempts according to passage answerability and role ability; (3) More complex scenarios with nested, multi-turn and prioritized instructions. The final RoleMRC features a 10.2k role profile meta-pool, 37.9k well-synthesized role-playing instructions, and 1.4k testing samples. We develop a pipeline to quantitatively evaluate the fine-grained role-playing and instruction-following capabilities of several mainstream LLMs, as well as models that are fine-tuned on our data. Moreover, cross-evaluation on external role-playing datasets confirms that models fine-tuned on RoleMRC enhances instruction-following without compromising general role-playing and reasoning capabilities. We also probe the neural-level activation maps of different capabilities over post-tuned LLMs. Access to our RoleMRC, RoleMRC-mix and Codes: https://github.com/LuJunru/RoleMRC.
Automated text readability prediction is widely used in many real-world scenarios. Over the past century, such measures have primarily been developed and evaluated on reading comprehension outcomes and on human annotations of text readability levels. In this work, we propose an alternative, eye tracking-based cognitive framework which directly taps into a key aspect of readability: reading ease. We use this framework for evaluating a broad range of prominent readability measures, including two systems widely used in education, by quantifying their ability to account for reading facilitation effects in text simplification, as well as text reading ease more broadly. Our analyses suggest that existing readability measures are poor predictors of reading facilitation and reading ease, outperformed by word properties commonly used in psycholinguistics, and in particular by surprisal.
English as a Second Language (ESL) learners often encounter unknown words that hinder their text comprehension. Automatically detecting these words as users read can enable computing systems to provide just-in-time definitions, synonyms, or contextual explanations, thereby helping users learn vocabulary in a natural and seamless manner. This paper presents EyeLingo, a transformer-based machine learning method that predicts the probability of unknown words based on text content and eye gaze trajectory in real time with high accuracy. A 20-participant user study revealed that our method can achieve an accuracy of 97.6%, and an F1-score of 71.1%. We implemented a real-time reading assistance prototype to show the effectiveness of EyeLingo. The user study shows improvement in willingness to use and usefulness compared to baseline methods.
We introduce an evaluation methodology for reading comprehension tasks based on the intuition that certain examples, by the virtue of their linguistic complexity, consistently yield lower scores regardless of model size or architecture. We capitalize on semantic frame annotation for characterizing this complexity, and study seven complexity factors that may account for model's difficulty. We first deploy this methodology on a carefully annotated French reading comprehension benchmark showing that two of those complexity factors are indeed good predictors of models' failure, while others are less so. We further deploy our methodology on a well studied English benchmark by using Chat-GPT as a proxy for semantic annotation. Our study reveals that fine-grained linguisticallymotivated automatic evaluation of a reading comprehension task is not only possible, but helps understand models' abilities to handle specific linguistic characteristics of input examples. It also shows that current state-of-the-art models fail with some for those characteristics which suggests that adequately handling them requires more than merely increasing model size.
Large language models (LLMs) have demonstrated remarkable progress in understanding long-context inputs. However, benchmarks for evaluating the long-context reasoning abilities of LLMs fall behind the pace. Existing benchmarks often focus on a narrow range of tasks or those that do not demand complex reasoning. To address this gap and enable a more comprehensive evaluation of the long-context reasoning capabilities of current LLMs, we propose a new synthetic benchmark, LongReason, which is constructed by synthesizing long-context reasoning questions from a varied set of short-context reasoning questions through context expansion. LongReason consists of 794 multiple-choice reasoning questions with diverse reasoning patterns across three task categories: reading comprehension, logical inference, and mathematical word problems. We evaluate 21 LLMs on LongReason, revealing that most models experience significant performance drops as context length increases. Our further analysis shows that even state-of-the-art LLMs still have significant room for improvement in providing robust reasoning across different tasks. We will open-source LongReason to support the comprehensive evaluation of LLMs' long-context reasoning capabilities.
Short-reading comprehension questions help students understand text structure but lack effective feedback. Students struggle to identify and correct errors, while manual feedback creation is labor-intensive. This highlights the need for automated feedback linking responses to a scoring rubric for deeper comprehension. Despite advances in Natural Language Processing (NLP), research has focused on automatic grading, with limited work on feedback generation. To address this, we propose a system that generates feedback for student responses. Our contributions are twofold. First, we introduce the first system for feedback on short-answer reading comprehension. These answers are derived from the text, requiring structural understanding. We propose an "answer diagnosis graph," integrating the text's logical structure with feedback templates. Using this graph and NLP techniques, we estimate students' comprehension and generate targeted feedback. Second, we evaluate our feedback through an experiment with Japanese high school students (n=39). They answered two 70-80 word questions and were divided into two groups with minimal academic differences. One received a model answer, the other system-generated feedback. Both re-answered the questions, and we compared score changes. A questionnaire assessed perceptions and motivation. Results showed no significant score improvement between groups, but system-generated feedback helped students identify errors and key points in the text. It also significantly increased motivation. However, further refinement is needed to enhance text structure understanding.
Detecting out-of-distribution (OOD) data is crucial in real-world machine learning applications, particularly in safety-critical domains. Existing methods often leverage language information from vision-language models (VLMs) to enhance OOD detection by improving confidence estimation through rich class-wise text information. However, when building OOD detection score upon on in-distribution (ID) text-image affinity, existing works either focus on each ID class or whole ID label sets, overlooking inherent ID classes' connection. We find that the semantic information across different ID classes is beneficial for effective OOD detection. We thus investigate the ability of image-text comprehension among different semantic-related ID labels in VLMs and propose a novel post-hoc strategy called SimLabel. SimLabel enhances the separability between ID and OOD samples by establishing a more robust image-class similarity metric that considers consistency over a set of similar class labels. Extensive experiments demonstrate the superior performance of SimLabel on various zero-shot OOD detection benchmarks. The proposed model is also extended to various VLM-backbones, demonstrating its good generalization ability. Our demonstration and implementation codes are available at: https://github.com/ShuZou-1/SimLabel.
The task of policy compliance detection (PCD) is to determine if a scenario is in compliance with respect to a set of written policies. In a conversational setting, the results of PCD can indicate if clarifying questions must be asked to determine compliance status. Existing approaches usually claim to have reasoning capabilities that are latent or require a large amount of annotated data. In this work, we propose logical decomposition for policy compliance (LDPC): a neuro-symbolic framework to detect policy compliance using large language models (LLMs) in a few-shot setting. By selecting only a few exemplars alongside recently developed prompting techniques, we demonstrate that our approach soundly reasons about policy compliance conversations by extracting sub-questions to be answered, assigning truth values from contextual information, and explicitly producing a set of logic statements from the given policies. The formulation of explicit logic graphs can in turn help answer PCDrelated questions with increased transparency and explainability. We apply this approach to the popular PCD and conversational machine reading benchmark, ShARC, and show competitive performance with no task-specific finetuning. We also leverage the inherently interpretable architecture of LDPC to understand where errors occur, revealing ambiguities in the ShARC dataset and highlighting the challenges involved with reasoning for conversational question answering.
Evaluating text comprehension in educational settings is critical for understanding student performance and improving curricular effectiveness. This study investigates the capability of state-of-the-art language models-RoBERTa Base, Bangla-BERT, and BERT Base-in automatically assessing Bangla passage-based question-answering from the National Curriculum and Textbook Board (NCTB) textbooks for classes 6-10. A dataset of approximately 3,000 Bangla passage-based question-answering instances was compiled, and the models were evaluated using F1 Score and Exact Match (EM) metrics across various hyperparameter configurations. Our findings revealed that Bangla-BERT consistently outperformed the other models, achieving the highest F1 (0.75) and EM (0.53) scores, particularly with smaller batch sizes, the inclusion of stop words, and a moderate learning rate. In contrast, RoBERTa Base demonstrated the weakest performance, with the lowest F1 (0.19) and EM (0.27) scores under certain configurations. The results underscore the importance of fine-tuning hyperparameters for optimizing model performance and highlight the potential of machine learning models in evaluating text comprehension in educational contexts. However, limitations such as dataset size, spelling inconsistencies, and computational constraints emphasize the need for further research to enhance the robustness and applicability of these models. This study lays the groundwork for the future development of automated evaluation systems in educational institutions, providing critical insights into model performance in the context of Bangla text comprehension.
Retrieval-Augmented Generation (RAG) systems have become pivotal in leveraging vast corpora to generate informed and contextually relevant responses, notably reducing hallucinations in Large Language Models. Despite significant advancements, these systems struggle to efficiently process and retrieve information from large datasets while maintaining a comprehensive understanding of the context. This paper introduces SKETCH, a novel methodology that enhances the RAG retrieval process by integrating semantic text retrieval with knowledge graphs, thereby merging structured and unstructured data for a more holistic comprehension. SKETCH, demonstrates substantial improvements in retrieval performance and maintains superior context integrity compared to traditional methods. Evaluated across four diverse datasets: QuALITY, QASPER, NarrativeQA, and Italian Cuisine-SKETCH consistently outperforms baseline approaches on key RAGAS metrics such as answer_relevancy, faithfulness, context_precision and context_recall. Notably, on the Italian Cuisine dataset, SKETCH achieved an answer relevancy of 0.94 and a context precision of 0.99, representing the highest performance across all evaluated metrics. These results highlight SKETCH's capability in delivering more accurate and contextually relevant responses, setting new benchmarks for future retrieval systems.
Why do we build local large language models (LLMs)? What should a local LLM learn from the target language? Which abilities can be transferred from other languages? Do language-specific scaling laws exist? To explore these research questions, we evaluated 35 Japanese, English, and multilingual LLMs on 19 evaluation benchmarks for Japanese and English, taking Japanese as a local language. Adopting an observational approach, we analyzed correlations of benchmark scores, and conducted principal component analysis (PCA) on the scores to derive \textit{ability factors} of local LLMs. We found that training on English text can improve the scores of academic subjects in Japanese (JMMLU). In addition, it is unnecessary to specifically train on Japanese text to enhance abilities for solving Japanese code generation, arithmetic reasoning, commonsense, and reading comprehension tasks. In contrast, training on Japanese text could improve question-answering tasks about Japanese knowledge and English-Japanese translation, which indicates that abilities for solving these two tasks can be regarded as \textit{Japanese abilities} for LLMs. Furthermore, we confirmed that the Japanese abilities scale with the computational budget for Japanese text.
The domain of Natural Language Processing (NLP) has experienced notable progress in the evolution of Bangla Question Answering (QA) systems. This paper presents a comprehensive review of seven research articles that contribute to the progress in this domain. These research studies explore different aspects of creating question-answering systems for the Bangla language. They cover areas like collecting data, preparing it for analysis, designing models, conducting experiments, and interpreting results. The papers introduce innovative methods like using LSTM-based models with attention mechanisms, context-based QA systems, and deep learning techniques based on prior knowledge. However, despite the progress made, several challenges remain, including the lack of well-annotated data, the absence of high-quality reading comprehension datasets, and difficulties in understanding the meaning of words in context. Bangla QA models' precision and applicability are constrained by these challenges. This review emphasizes the significance of these research contributions by highlighting the developments achieved in creating Bangla QA systems as well as the ongoing effort required to get past roadblocks and improve the performance of these systems for actual language comprehension tasks.
Recent studies have shown that generative language models often reflect and amplify societal biases in their outputs. However, these studies frequently conflate observed biases with other task-specific shortcomings, such as comprehension failure. For example, when a model misinterprets a text and produces a response that reinforces a stereotype, it becomes difficult to determine whether the issue arises from inherent bias or from a misunderstanding of the given content. In this paper, we conduct a multi-faceted evaluation that distinctly disentangles bias from flaws within the reading comprehension task. We propose a targeted stereotype mitigation framework that implicitly mitigates observed stereotypes in generative models through instruction-tuning on general-purpose datasets. We reduce stereotypical outputs by over 60% across multiple dimensions -- including nationality, age, gender, disability, and physical appearance -- by addressing comprehension-based failures, and without relying on explicit debiasing techniques. We evaluate several state-of-the-art generative models to demonstrate the effectiveness of our approach while maintaining the overall utility. Our findings highlight the need to critically disentangle the concept of `bias' from other types of errors to build more targeted and effective mitigation strategies. CONTENT WARNING: Some examples contain offensive stereotypes.
This study examines whether large language models (LLMs), such as ChatGPT, specifically the latest GPT-4o-mini, exhibit sensitivity to repeated prompts and whether repeating a question can improve response accuracy. We hypothesize that reiterating a question within a single prompt might enhance the model's focus on key elements of the query. To test this, we evaluate ChatGPT's performance on a large sample of two reading comprehension datasets under both open-book and closed-book settings, varying the repetition of each question to 1, 3, or 5 times per prompt. Our findings indicate that the model does not demonstrate sensitivity to repeated questions, highlighting its robustness and consistency in this context.
The purpose of this work is to share an English-Yor\`ub\'a evaluation dataset for open-book reading comprehension and text generation to assess the performance of models both in a high- and a low- resource language. The dataset contains 358 questions and answers on 338 English documents and 208 Yor\`ub\'a documents. The average document length is ~ 10k words for English and 430 words for Yor\`ub\'a. Experiments show a consistent disparity in performance between the two languages, with Yor\`ub\'a falling behind English for automatic metrics even if documents are much shorter for this language. For a small set of documents with comparable length, performance of Yor\`ub\'a drops by x2.5 times. When analyzing performance by length, we observe that Yor\`ub\'a decreases performance dramatically for documents that reach 1500 words while English performance is barely affected at that length. Our dataset opens the door to showcasing if English LLM reading comprehension capabilities extend to Yor\`ub\'a, which for the evaluated LLMs is not the case.
We introduce the first highly multilingual speech and American Sign Language (ASL) comprehension dataset by extending BELEBELE. Our dataset covers 74 spoken languages at the intersection of BELEBELE and FLEURS, and one sign language (ASL). We evaluate 2M-BELEBELE dataset for both 5-shot and zero-shot settings and across languages, the speech comprehension accuracy is ~ 2-3% average lower compared to reading comprehension.
Text serves as the key control signal in video generation due to its narrative nature. To render text descriptions into video clips, current video diffusion models borrow features from text encoders yet struggle with limited text comprehension. The recent success of large language models (LLMs) showcases the power of decoder-only transformers, which offers three clear benefits for text-to-video (T2V) generation, namely, precise text understanding resulting from the superior scalability, imagination beyond the input text enabled by next token prediction, and flexibility to prioritize user interests through instruction tuning. Nevertheless, the feature distribution gap emerging from the two different text modeling paradigms hinders the direct use of LLMs in established T2V models. This work addresses this challenge with Mimir, an end-to-end training framework featuring a carefully tailored token fuser to harmonize the outputs from text encoders and LLMs. Such a design allows the T2V model to fully leverage learned video priors while capitalizing on the text-related capability of LLMs. Extensive quantitative and qualitative results demonstrate the effectiveness of Mimir in generating high-quality videos with excellent text comprehension, especially when processing short captions and managing shifting motions. Project page: https://lucaria-academy.github.io/Mimir/
Relevance modeling between queries and items stands as a pivotal component in commercial search engines, directly affecting the user experience. Given the remarkable achievements of large language models (LLMs) in various natural language processing (NLP) tasks, LLM-based relevance modeling is gradually being adopted within industrial search systems. Nevertheless, foundational LLMs lack domain-specific knowledge and do not fully exploit the potential of in-context learning. Furthermore, structured item text remains underutilized, and there is a shortage in the supply of corresponding queries and background knowledge. We thereby propose CPRM (Continual Pre-training for Relevance Modeling), a framework designed for the continual pre-training of LLMs to address these issues. Our CPRM framework includes three modules: 1) employing both queries and multi-field item to jointly pre-train for enhancing domain knowledge, 2) applying in-context pre-training, a novel approach where LLMs are pre-trained on a sequence of related queries or items, and 3) conducting reading comprehension on items to produce associated domain knowledge and background information (e.g., generating summaries and corresponding queries) to further strengthen LLMs. Results on offline experiments and online A/B testing demonstrate that our model achieves convincing performance compared to strong baselines.
Large Language Models (LLMs), with gradually improving reading comprehension and reasoning capabilities, are being applied to a range of complex language tasks, including the automatic generation of language data for various purposes. However, research on applying LLMs for automatic data generation in low-resource languages like Vietnamese is still underdeveloped and lacks comprehensive evaluation. In this paper, we explore the use of LLMs for automatic data generation for the Vietnamese fact-checking task, which faces significant data limitations. Specifically, we focus on fact-checking data where claims are synthesized from multiple evidence sentences to assess the information synthesis capabilities of LLMs. We develop an automatic data construction process using simple prompt techniques on LLMs and explore several methods to improve the quality of the generated data. To evaluate the quality of the data generated by LLMs, we conduct both manual quality assessments and performance evaluations using language models. Experimental results and manual evaluations illustrate that while the quality of the generated data has significantly improved through fine-tuning techniques, LLMs still cannot match the data quality produced by humans.
Scientific literature is typically dense, requiring significant background knowledge and deep comprehension for effective engagement. We introduce SciDQA, a new dataset for reading comprehension that challenges LLMs for a deep understanding of scientific articles, consisting of 2,937 QA pairs. Unlike other scientific QA datasets, SciDQA sources questions from peer reviews by domain experts and answers by paper authors, ensuring a thorough examination of the literature. We enhance the dataset's quality through a process that carefully filters out lower quality questions, decontextualizes the content, tracks the source document across different versions, and incorporates a bibliography for multi-document question-answering. Questions in SciDQA necessitate reasoning across figures, tables, equations, appendices, and supplementary materials, and require multi-document reasoning. We evaluate several open-source and proprietary LLMs across various configurations to explore their capabilities in generating relevant and factual responses. Our comprehensive evaluation, based on metrics for surface-level similarity and LLM judgements, highlights notable performance discrepancies. SciDQA represents a rigorously curated, naturally derived scientific QA dataset, designed to facilitate research on complex scientific text understanding.