The increasing demand for medical imaging has surpassed the capacity of available radiologists, leading to diagnostic delays and potential misdiagnoses. Artificial intelligence (AI) techniques, particularly in automatic medical report generation (AMRG), offer a promising solution to this dilemma. This review comprehensively examines AMRG methods from 2021 to 2024. It (i) presents solutions to primary challenges in this field, (ii) explores AMRG applications across various imaging modalities, (iii) introduces publicly available datasets, (iv) outlines evaluation metrics, (v) identifies techniques that significantly enhance model performance, and (vi) discusses unresolved issues and potential future research directions. This paper aims to provide a comprehensive understanding of the existing literature and inspire valuable future research.

The guidance from capability evaluations has greatly propelled the progress of both human society and Artificial Intelligence. However, as LLMs evolve, it becomes challenging to construct evaluation benchmarks for them with accurate labels on hard tasks that approach the boundaries of human capabilities. To credibly conduct evaluation without accurate labels (denoted as poor-supervised evaluation), we propose the PoEM framework. We first prove that the capability of a model can be equivalently assessed by the consistency between it and certain reference model, when their prediction distributions are independent and the sample size is infinite. To alleviate the insufficiencies of the conditions in reality, we further introduce an algorithm that treats humans (when available) and the models under evaluation as reference models, alternately conducting model weights calibration and filtering during E-step and M-step. Comprehensive experiments across 3 types of tasks with 16 mainstream LLMs have shown that PoEM under poor supervision can achieve an average of 0.98 Pearson correlation coefficient with supervised evaluation results, demonstrating good effectiveness, efficiency and generalizability. More generally, PoEM has advanced the evaluation paradigm evolution from human-centric to human&model-centric by treating both of them as reference models, mitigating the limitations of human evaluation in the era of LLMs.

In-Context Learning (ICL) enables large language models (LLMs) to achieve rapid task adaptation by learning from demonstrations. With the increase in available context length of LLMs, recent experiments have shown that the performance of ICL does not necessarily scale well in many-shot (demonstration) settings. We theoretically and experimentally confirm that the reason lies in more demonstrations dispersing the model attention from the query, hindering its understanding of key content. Inspired by how humans learn from examples, we propose a training-free method FocusICL, which conducts triviality filtering to avoid attention being diverted by unimportant contents at token-level and operates hierarchical attention to further ensure sufficient attention towards current query at demonstration-level. We also design an efficient hyperparameter searching strategy for FocusICL based on model perplexity of demonstrations. Comprehensive experiments validate that FocusICL achieves an average performance improvement of 5.2% over vanilla ICL and scales well with many-shot demonstrations.

Any report frames issues to favor a particular interpretation by highlighting or excluding certain aspects of a story. Despite the widespread use of framing in disinformation, framing properties and detection methods remain underexplored outside the English-speaking world. We explore how multilingual framing of the same issue differs systematically. We use eight years of Russia-backed disinformation campaigns, spanning 8k news articles in 4 languages targeting 15 countries. We find that disinformation campaigns consistently and intentionally favor specific framing, depending on the target language of the audience. We further discover how Russian-language articles consistently highlight selected frames depending on the region of the media coverage. We find that the two most prominent models for automatic frame analysis underperform and show high disagreement, highlighting the need for further research.

Many languages have vast amounts of handwritten texts, such as ancient scripts about folktale stories and historical narratives or contemporary documents and letters. Digitization of those texts has various applications, such as daily tasks, cultural studies, and historical research. Syriac is an ancient, endangered, and low-resourced language that has not received the attention it requires and deserves. This paper reports on a research project aimed at developing a optical character recognition (OCR) model based on the handwritten Syriac texts as a starting point to build more digital services for this endangered language. A dataset was created, KHAMIS (inspired by the East Syriac poet, Khamis bar Qardahe), which consists of handwritten sentences in the East Syriac script. We used it to fine-tune the Tesseract-OCR engine's pretrained Syriac model on handwritten data. The data was collected from volunteers capable of reading and writing in the language to create KHAMIS. KHAMIS currently consists of 624 handwritten Syriac sentences collected from 31 university students and one professor, and it will be partially available online and the whole dataset available in the near future for development and research purposes. As a result, the handwritten OCR model was able to achieve a character error rate of 1.097-1.610% and 8.963-10.490% on both training and evaluation sets, respectively, and both a character error rate of 18.89-19.71% and a word error rate of 62.83-65.42% when evaluated on the test set, which is twice as better than the default Syriac model of Tesseract.

Sign language translation has historically been peripheral to mainstream machine translation research. In order to help converge the fields, we introduce FLEURS-ASL, an extension of the multiway parallel benchmarks FLORES (for text) and FLEURS (for speech) to support their first sign language (as video), American Sign Language, translated by 5 Certified Deaf Interpreters. FLEURS-ASL can be used to evaluate a variety of tasks -- primarily sentence- and discourse-level translation -- between ASL and 200 other languages as text, or 102 languages as speech. We provide baselines for tasks from ASL to English text using a unified modeling approach that incorporates timestamp tokens and previous text tokens in a 34-second context window, trained on random video clips from YouTube-ASL. This model meets or exceeds the performance of phrase-level baselines while supporting a multitude of new tasks. We also use FLEURS-ASL to show that multimodal frontier models have virtually no understanding of ASL, underscoring the importance of including sign languages in standard evaluation suites.

Biases inherent in human endeavors pose significant challenges for machine learning, particularly in supervised learning that relies on potentially biased "ground truth" data. This reliance, coupled with models' tendency to generalize based on statistical maximal likelihood, can propagate and amplify biases, exacerbating societal issues. To address this, our study proposes a reflective methodology utilizing multiple Large Language Models (LLMs) engaged in a dynamic dialogue to uncover diverse perspectives. By leveraging conditional statistics, information theory, and divergence metrics, this novel approach fosters context-dependent linguistic behaviors, promoting unbiased outputs. Furthermore, it enables measurable progress tracking and explainable remediation actions to address identified biases.

Machine Translation of Culture-Specific Items (CSIs) poses significant challenges. Recent work on CSI translation has shown some success using Large Language Models (LLMs) to adapt to different languages and cultures; however, a deeper analysis is needed to examine the benefits and pitfalls of each method. In this paper, we introduce the ChineseMenuCSI dataset, the largest for Chinese-English menu corpora, annotated with CSI vs Non-CSI labels and a fine-grained test set. We define three levels of CSI figurativeness for a more nuanced analysis and develop a novel methodology for automatic CSI identification, which outperforms GPT-based prompts in most categories. Importantly, we are the first to integrate human translation theories into LLM-driven translation processes, significantly improving translation accuracy, with COMET scores increasing by up to 7 points.

Self-consistency (SC), a widely used decoding strategy for chain-of-thought reasoning, shows significant gains across various multi-step reasoning tasks but comes with a high cost due to multiple sampling with the preset size. Its variants, Adaptive self-consistency (ASC) and Early-stopping self-consistency (ESC), dynamically adjust the number of samples based on the posterior distribution of a set of pre-samples, reducing the cost of SC with minimal impact on performance. Both methods, however, do not exploit the prior information about question difficulty. It often results in unnecessary repeated sampling for easy questions that could be accurately answered with just one attempt, wasting resources. To tackle this problem, we propose Difficulty-Adaptive Self-Consistency (DSC), which leverages the difficulty information from both prior and posterior perspectives to adaptively allocate inference resources, further reducing the cost of SC. To demonstrate the effectiveness of DSC, we conduct extensive experiments on three popular categories of reasoning tasks: arithmetic, commonsense and symbolic reasoning on six benchmarks. The empirical results show that DSC consistently surpasses the strong baseline ASC and ESC in terms of costs by a significant margin, while attaining comparable performances.

Theoretical studies on the representation power of GNNs have been centered around understanding the equivalence of GNNs, using WL-Tests for detecting graph isomorphism. In this paper, we argue that such equivalence ignores the accompanying optimization issues and does not provide a holistic view of the GNN learning process. We illustrate these gaps between representation and optimization with examples and experiments. We also explore the existence of an implicit inductive bias (e.g. fully connected networks prefer to learn low frequency functions in their input space) in GNNs, in the context of graph classification tasks. We further prove theoretically that the message-passing layers in the graph, have a tendency to search for either discriminative subgraphs, or a collection of discriminative nodes dispersed across the graph, depending on the different global pooling layers used. We empirically verify this bias through experiments over real-world and synthetic datasets. Finally, we show how our work can help in incorporating domain knowledge via attention based architectures, and can evince their capability to discriminate coherent subgraphs.