This study compares automated and human scores with expert/backread scores for writing, reading, and science tasks in a K-12 assessment program. Analyses examine discrepancies by score level and reused prompts across administrations. Automated scores showed smaller discrepancies for writing but larger discrepancies for short-answer tasks; reused prompts revealed localized shifts.

Pricing automation in large-scale tourism is challenging because travel orders are highly unstructured, while pricing policies are complex, rapidly evolving, and inherently open-ended. Traditional rule engines are brittle and costly to maintain, whereas unconstrained LLM agents lack the reliability and auditability required for financial decisions. We present a production-grade LLM-powered pricing system with a strict decision boundary: LLMs perform structured extraction and bounded policy/path selection, while all numeric pricing, including total-price computation, is executed deterministically. Policies are compiled into interpretable condition trees, enabling open-ended support for new clauses and evolving rules without code changes, while exposing auditable artifacts for human-in-the-loop control. Periodic fine-tuning on logged traces further improves tree induction and path matching. Deployed at a municipal state-owned tourism enterprise across 7 scenic sites and 12 business categories with 1,500+ operators and 1,000+ active policies, the system processed 3,960 orders in six months, reduced the order management team from 15-20 to 3, and cut per-order handling time from 10 minutes to <2 minutes.

Recent advancements in large language models (LLMs) have showcased impressive code generation capabilities, primarily evaluated through language-to-code benchmarks. However, these benchmarks may not fully capture a model’s code understanding abilities. We introduce CodeJudge-Eval (CJ-Eval), a novel benchmark designed to assess LLMs’ code understanding abilities from the perspective of code judging rather than code generation. CJ-Eval challenges models to determine the correctness of provided code solutions, encompassing various error types and compilation issues. By leveraging a diverse set of problems and a fine-grained judging system, CJ-Eval addresses the limitations of traditional benchmarks, including the potential memorization of solutions. Evaluation of 12 well-known LLMs on CJ-Eval reveals that even state-of-the-art models struggle, highlighting the benchmark’s ability to probe deeper into models’ code understanding abilities. Our benchmark is available at https://github.com/CodeLLM-Research/CodeJudge-Eval .

With the pervasive use of large language models (LLMs) across various domains, addressing the inherent ableist biases within these models requires more attention and resolution. This paper examines ableism in three LLMs (GPT-3.5, GPT-4, and Llama 3) by analyzing the intersection of disability with two additional social categories: gender and social class. Utilizing two task-specific prompts, we generated and analyzed text outputs with two metrics, VADER and regard, to evaluate sentiment and social perception biases within the responses. Our results indicate a marked improvement in bias mitigation from GPT-3.5 to GPT-4, with the latter demonstrating more positive sentiments overall, while Llama 3 showed comparatively weaker performance. Additionally, our findings underscore the complexity of intersectional biases: These biases are shaped by the combined effects of disability, gender, and class, which alter the expression and perception of ableism in LLM outputs. This research highlights the necessity for more nuanced and inclusive bias mitigation strategies in AI development, contributing to the ongoing dialogue on ethical AI practices.

Social media abounds with multimodal sarcasm, and identifying sarcasm targets is particularly challenging due to the implicit incongruity not directly evident in the text and image modalities. Current methods for Multimodal Sarcasm Target Identification (MSTI) predominantly focus on superficial indicators in an end-to-end manner, overlooking the nuanced understanding of multimodal sarcasm conveyed through both the text and image. This paper proposes a versatile MSTI framework with a coarse-to-fine paradigm, by augmenting sarcasm explainability with reasoning and pre-training knowledge. Inspired by the powerful capacity of Large Multimodal Models (LMMs) on multimodal reasoning, we first engage LMMs to generate competing rationales for coarser-grained pre-training of a small language model on multimodal sarcasm detection. We then propose fine-tuning the model for finer-grained sarcasm target identification. Our framework is thus empowered to adeptly unveil the intricate targets within multimodal sarcasm and mitigate the negative impact posed by potential noise inherently in LMMs. Experimental results demonstrate that our model far outperforms state-of-the-art MSTI methods, and markedly exhibits explainability in deciphering sarcasm as well.

The age of social media is rife with memes. Understanding and detecting harmful memes pose a significant challenge due to their implicit meaning that is not explicitly conveyed through the surface text and image. However, existing harmful meme detection approaches only recognize superficial harm-indicative signals in an end-to-end classification manner but ignore in-depth cognition of the meme text and image. In this paper, we attempt to detect harmful memes based on advanced reasoning over the interplay of multimodal information in memes. Inspired by the success of Large Language Models (LLMs) on complex reasoning, we first conduct abductive reasoning with LLMs. Then we propose a novel generative framework to learn reasonable thoughts from LLMs for better multimodal fusion and lightweight fine-tuning, which consists of two training stages: 1) Distill multimodal reasoning knowledge from LLMs; and 2) Fine-tune the generative framework to infer harmfulness. Extensive experiments conducted on three meme datasets demonstrate that our proposed approach achieves superior performance than state-of-the-art methods on the harmful meme detection task.

Social media has not only facilitated news consumption, but also led to the wide spread of fake news. Because news articles in social media is usually condensed and full of knowledge entities, existing methods of fake news detection use external entity knowledge. However, majority of these methods focus on news entity information and ignore the structured knowledge among news entities. To address this issue, in this work, we propose a Knowledge grAPh enhAnced Language Model (KAPALM) which is a novel model that fuses coarse- and fine-grained representations of entity knowledge from Knowledge Graphs (KGs). Firstly, we identify entities in news content and link them to entities in KGs. Then, a subgraph of KGs is extracted to provide structured knowledge of entities in KGs and fed into a graph neural network to obtain the coarse-grained knowledge representation. This subgraph is pruned to provide fine-grained knowledge and fed into the attentive graph and graph pooling layer. Finally, we integrate the coarse- and fine-grained entity knowledge representations with the textual representation for fake news detection. The experimental results on two benchmark datasets show that our method is superior to state-of-the-art baselines. In addition, it is competitive in the few-shot scenario.

Fake news debunking primarily focuses on determining the truthfulness of news articles, which oversimplifies the issue as fake news often combines elements of both truth and falsehood. Thus, it becomes crucial to identify specific instances of misinformation within the articles. In this research, we investigate a novel task in the field of fake news debunking, which involves detecting sentence-level misinformation. One of the major challenges in this task is the absence of a training dataset with sentence-level annotations regarding veracity. Inspired by the Multiple Instance Learning (MIL) approach, we propose a model called Weakly Supervised Detection of Misinforming Sentences (WSDMS). This model only requires bag-level labels for training but is capable of inferring both sentence-level misinformation and article-level veracity, aided by relevant social media conversations that are attentively contextualized with news sentences. We evaluate WSDMS on three real-world benchmarks and demonstrate that it outperforms existing state-of-the-art baselines in debunking fake news at both the sentence and article levels.

Image-text retrieval is a fundamental cross-modal task that takes image/text as a query to retrieve relevant data of another type. The large-scale two-stream pre-trained models like CLIP have achieved tremendous success in this area. They embed the images and texts into instance representations with two separate encoders, aligning them on the instance-level with contrastive learning. Beyond this, the following works adopt the fine-grained token-level interaction (Masked Language and Image Modeling) to boost performance further. However, the vanilla token-level objectives are not designed to aggregate the image-text alignment information into the instance representations, but the token representations, causing a gap between pre-training and application. To address this issue, we carefully design two novel conditioned token-level pre-training objectives, Conditioned Masked Language and Image Modeling (ConMLM and ConMIM), forcing models to aggregate the token-level alignment information into the instance representations. Combing with the instance-level contrastive learning, we propose our cross-modal dense retrieval framework, Conditioned Language-Image Pre-training (ConLIP). Experimental results on two popular cross-modal retrieval benchmarks (MSCOCO and Flickr30k) reveal the effectiveness of our methods.

Existing fake news detection methods aim to classify a piece of news as true or false and provide veracity explanations, achieving remarkable performances. However, they often tailor automated solutions on manual fact-checked reports, suffering from limited news coverage and debunking delays. When a piece of news has not yet been fact-checked or debunked, certain amounts of relevant raw reports are usually disseminated on various media outlets, containing the wisdom of crowds to verify the news claim and explain its verdict. In this paper, we propose a novel Coarse-to-fine Cascaded Evidence-Distillation (CofCED) neural network for explainable fake news detection based on such raw reports, alleviating the dependency on fact-checked ones. Specifically, we first utilize a hierarchical encoder for web text representation, and then develop two cascaded selectors to select the most explainable sentences for verdicts on top of the selected top-K reports in a coarse-to-fine manner. Besides, we construct two explainable fake news datasets, which is publicly available. Experimental results demonstrate that our model significantly outperforms state-of-the-art detection baselines and generates high-quality explanations from diverse evaluation perspectives.

Nested Named Entity Recognition (NNER) has been extensively studied, aiming to identify all nested entities from potential spans (i.e., one or more continuous tokens). However, recent studies for NNER either focus on tedious tagging schemas or utilize complex structures, which fail to learn effective span representations from the input sentence with highly nested entities. Intuitively, explicit span representations will contribute to NNER due to the rich context information they contain. In this study, we propose a Hierarchical Transformer (HiTRANS) network for the NNER task, which decomposes the input sentence into multi-grained spans and enhances the representation learning in a hierarchical manner. Specifically, we first utilize a two-phase module to generate span representations by aggregating context information based on a bottom-up and top-down transformer network. Then a label prediction layer is designed to recognize nested entities hierarchically, which naturally explores semantic dependencies among different spans. Experiments on GENIA, ACE-2004, ACE-2005 and NNE datasets demonstrate that our proposed method achieves much better performance than the state-of-the-art approaches.

Rumors are rampant in the era of social media. Conversation structures provide valuable clues to differentiate between real and fake claims. However, existing rumor detection methods are either limited to the strict relation of user responses or oversimplify the conversation structure. In this study, to substantially reinforces the interaction of user opinions while alleviating the negative impact imposed by irrelevant posts, we first represent the conversation thread as an undirected interaction graph. We then present a Claim-guided Hierarchical Graph Attention Network for rumor classification, which enhances the representation learning for responsive posts considering the entire social contexts and attends over the posts that can semantically infer the target claim. Extensive experiments on three Twitter datasets demonstrate that our rumor detection method achieves much better performance than state-of-the-art methods and exhibits a superior capacity for detecting rumors at early stages.

This paper describes our contribution to SemEval 2020 Task 8: Memotion Analysis. Our system learns multi-modal embeddings from text and images in order to classify Internet memes by sentiment. Our model learns text embeddings using BERT and extracts features from images with DenseNet, subsequently combining both features through concatenation. We also compare our results with those produced by DenseNet, ResNet, BERT, and BERT-ResNet. Our results show that image classification models have the potential to help classifying memes, with DenseNet outperforming ResNet. Adding text features is however not always helpful for Memotion Analysis.

Product-related question answering platforms nowadays are widely employed in many E-commerce sites, providing a convenient way for potential customers to address their concerns during online shopping. However, the misinformation in the answers on those platforms poses unprecedented challenges for users to obtain reliable and truthful product information, which may even cause a commercial loss in E-commerce business. To tackle this issue, we investigate to predict the veracity of answers in this paper and introduce AnswerFact, a large scale fact checking dataset from product question answering forums. Each answer is accompanied by its veracity label and associated evidence sentences, providing a valuable testbed for evidence-based fact checking tasks in QA settings. We further propose a novel neural model with tailored evidence ranking components to handle the concerned answer veracity prediction problem. Extensive experiments are conducted with our proposed model and various existing fact checking methods, showing that our method outperforms all baselines on this task.

Rumors are manufactured with no respect for accuracy, but can circulate quickly and widely by “word-of-post” through social media conversations. Conversation tree encodes important information indicative of the credibility of rumor. Existing conversation-based techniques for rumor detection either just strictly follow tree edges or treat all the posts fully-connected during feature learning. In this paper, we propose a novel detection model based on tree transformer to better utilize user interactions in the dialogue where post-level self-attention plays the key role for aggregating the intra-/inter-subtree stances. Experimental results on the TWITTER and PHEME datasets show that the proposed approach consistently improves rumor detection performance.

Claim verification is generally a task of verifying the veracity of a given claim, which is critical to many downstream applications. It is cumbersome and inefficient for human fact-checkers to find consistent pieces of evidence, from which solid verdict could be inferred against the claim. In this paper, we propose a novel end-to-end hierarchical attention network focusing on learning to represent coherent evidence as well as their semantic relatedness with the claim. Our model consists of three main components: 1) A coherence-based attention layer embeds coherent evidence considering the claim and sentences from relevant articles; 2) An entailment-based attention layer attends on sentences that can semantically infer the claim on top of the first attention; and 3) An output layer predicts the verdict based on the embedded evidence. Experimental results on three public benchmark datasets show that our proposed model outperforms a set of state-of-the-art baselines.

Automatic rumor detection is technically very challenging. In this work, we try to learn discriminative features from tweets content by following their non-sequential propagation structure and generate more powerful representations for identifying different type of rumors. We propose two recursive neural models based on a bottom-up and a top-down tree-structured neural networks for rumor representation learning and classification, which naturally conform to the propagation layout of tweets. Results on two public Twitter datasets demonstrate that our recursive neural models 1) achieve much better performance than state-of-the-art approaches; 2) demonstrate superior capacity on detecting rumors at very early stage.

We describe our system for participating in SemEval-2017 Task 3 on Community Question Answering. Our approach relies on combining a rich set of various types of features: semantic and metadata. The most important group turned out to be the metadata feature and the semantic vectors trained on QatarLiving data. In the main Subtask C, our primary submission was ranked fourth, with a MAP of 13.48 and accuracy of 97.08. In Subtask A, our primary submission get into the top 50%.

How fake news goes viral via social media? How does its propagation pattern differ from real stories? In this paper, we attempt to address the problem of identifying rumors, i.e., fake information, out of microblog posts based on their propagation structure. We firstly model microblog posts diffusion with propagation trees, which provide valuable clues on how an original message is transmitted and developed over time. We then propose a kernel-based method called Propagation Tree Kernel, which captures high-order patterns differentiating different types of rumors by evaluating the similarities between their propagation tree structures. Experimental results on two real-world datasets demonstrate that the proposed kernel-based approach can detect rumors more quickly and accurately than state-of-the-art rumor detection models.