Table Representation Learning (TRL) models are commonly pre-trained on large open-domain datasets comprising millions of tables and then used to address downstream tasks. Choosing the right TRL model to use on proprietary data can be challenging, as the best results depend on the content domain, schema, and data quality. Our purpose is to support end-users in testing TRL models on proprietary data in two established SQL-centric tasks, i.e., Question Answering (QA) and Semantic Parsing (SP). We present QATCH (Query-Aided TRLChecklist), a toolbox to highlight TRL models' strengths and weaknesses on relational tables unseen at training time. For an input table, QATCH automatically generates a testing checklist tailored to QA and SP. Checklist generation is driven by a SQL query engine that crafts tests of different complexity. This design facilitates inherent portability, allowing the checks to be used by alternative models. We also introduce a set of cross-task performance metrics evaluating the TRL model's performance over its output. Finally, we show how QATCH automatically generates tests for proprietary datasets to evaluate various state-of-the-art models including TAPAS, TAPEX, and CHATGPT.

We measure LLMs' output error at pairwise text comparison, noting the probability of error in their preferences. Our method does not rely on the ground truth and supports two scenarios: (i) uniform error rate regardless of the order of comparison, estimated with two comparisons for each text pair with either text placed first; (ii) binary positional bias assuming distinct error rates for the two orders of comparison, estimated with repeated comparisons between the texts. The Copeland counting constructs a ranking over the compared texts from pairwise preferences; the ranking reveals the poor scalability of LLM-based pairwise comparison and helps yield the estimates for LLMs' error rates. We apply the method to six LLMs (ChatGPT, Claude, DeepSeek, Gemini, Grok, Qwen) with five types of text input and obtain consistent estimates of LLMs' error. In general, the measured two positional bias terms are similar, close to the uniform error. Considering both the error rates and the robustness to the variation of prompts, Claude obtained the most desirable performance in this experiment. Our model outperforms the biased Bradley-Terry model and the commutativity score in indicating LLMs' error at this task.

Recently, the powerful large language models (LLMs) have been instrumental in propelling the progress of recommender systems (RS). However, while these systems have flourished, their susceptibility to security threats has been largely overlooked. In this work, we reveal that the introduction of LLMs into recommendation models presents new security vulnerabilities due to their emphasis on the textual content of items. We demonstrate that attackers can significantly boost an item's exposure by merely altering its textual content during the testing phase, without requiring direct interference with the model's training process. Additionally, the attack is notably stealthy, as it does not affect the overall recommendation performance and the modifications to the text are subtle, making it difficult for users and platforms to detect. Our comprehensive experiments across four mainstream LLM-based recommendation models demonstrate the superior efficacy and stealthiness of our approach. Our work unveils a significant security gap in LLM-based recommendation systems and paves the way for future research on protecting these systems.

The diversity of knowledge encoded in large language models (LLMs) and their ability to apply this knowledge zero-shot in a range of settings makes them a promising candidate for use in decision-making. However, they are currently limited by their inability to reliably provide outputs which are explainable and contestable. In this paper, we attempt to reconcile these strengths and weaknesses by introducing a method for supplementing LLMs with argumentative reasoning. Concretely, we introduce argumentative LLMs, a method utilising LLMs to construct argumentation frameworks, which then serve as the basis for formal reasoning in decision-making. The interpretable nature of these argumentation frameworks and formal reasoning means that any decision made by the supplemented LLM may be naturally explained to, and contested by, humans. We demonstrate the effectiveness of argumentative LLMs experimentally in the decision-making task of claim verification. We obtain results that are competitive with, and in some cases surpass, comparable state-of-the-art techniques.

While text summarization is a well-known NLP task, in this paper, we introduce a novel and useful variant of it called functionality extraction from Git README files. Though this task is a text2text generation at an abstract level, it involves its own peculiarities and challenges making existing text2text generation systems not very useful. The motivation behind this task stems from a recent surge in research and development activities around the use of large language models for code-related tasks, such as code refactoring, code summarization, etc. We also release a human-annotated dataset called FuncRead, and develop a battery of models for the task. Our exhaustive experimentation shows that small size fine-tuned models beat any baseline models that can be designed using popular black-box or white-box large language models (LLMs) such as ChatGPT and Bard. Our best fine-tuned 7 Billion CodeLlama model exhibit 70% and 20% gain on the F1 score against ChatGPT and Bard respectively.

Background: The emergence of generative AI tools, empowered by Large Language Models (LLMs), has shown powerful capabilities in generating content. To date, the assessment of the usefulness of such content, generated by what is known as prompt engineering, has become an interesting research question. Objectives Using the mean of prompt engineering, we assess the similarity and closeness of such contents to real literature produced by scientists. Methods In this exploratory analysis, (1) we prompt-engineer ChatGPT and Google Bard to generate clinical content to be compared with literature counterparts, (2) we assess the similarities of the contents generated by comparing them with counterparts from biomedical literature. Our approach is to use text-mining approaches to compare documents and associated bigrams and to use network analysis to assess the terms' centrality. Results The experiments demonstrated that ChatGPT outperformed Google Bard in cosine document similarity (38% to 34%), Jaccard document similarity (23% to 19%), TF-IDF bigram similarity (47% to 41%), and term network centrality (degree and closeness). We also found new links that emerged in ChatGPT bigram networks that did not exist in literature bigram networks. Conclusions: The obtained similarity results show that ChatGPT outperformed Google Bard in document similarity, bigrams, and degree and closeness centrality. We also observed that ChatGPT offers linkage to terms that are connected in the literature. Such connections could inspire asking interesting questions and generate new hypotheses.

Hallucination plagues even frontier LLMs-- but how bad is it really for summarizing academic papers? We evaluate Factored Verification, a simple automated method for detecting hallucinations in abstractive summaries. This method sets a new SotA on hallucination detection in the summarization task of the HaluEval benchmark, achieving 76.2% accuracy. We then use this method to estimate how often language models hallucinate when summarizing across multiple academic papers and find 0.62 hallucinations in the average ChatGPT (16k) summary, 0.84 for GPT-4, and 1.55 for Claude 2. We ask models to self-correct using factored critiques and find that this lowers the number of hallucinations to 0.49 for ChatGPT, 0.46 for GPT-4, and 0.95 for Claude 2. The hallucinations we find are often subtle, so we advise caution when using models to synthesize academic papers.