From Calculation to Adjudication: Examining LLM judges on Mathematical Reasoning Tasks

To reduce the need for human annotations, large language models (LLMs) have been proposed as judges of the quality of other candidate models. LLM judges are typically evaluated by measuring the correlation with human judgments on generation tasks such as summarization or machine translation. In contrast, we study LLM judges on mathematical reasoning tasks. These tasks require multi-step reasoning, and the correctness of their solutions is verifiable, enabling a more objective evaluation. We perform a detailed performance analysis and find that the used judges are mostly unable to improve task performance but are able to pick the better model. Our analysis uncovers a strong correlation between judgment performance and the candidate model task performance. We observe that judges tend to choose the model of higher quality even if its answer is incorrect. Further, we show that it is possible to use statistics, such as the task performances of the individual models, to predict judgment performance. In an ablation, we either swap or mask the candidate answers and observe that judges often keep the original judgment, providing evidence that judges incorporate writing style in their judgments. In summary, we find that regularities in the judgments are quantifiable using statistical measures and provide various angles on exploiting them.

An Evaluation of Explanation Methods for Black-Box Detectors of Machine-Generated Text

The increasing difficulty to distinguish language-model-generated from human-written text has led to the development of detectors of machine-generated text (MGT). However, in many contexts, a black-box prediction is not sufficient, it is equally important to know on what grounds a detector made that prediction. Explanation methods that estimate feature importance promise to provide indications of which parts of an input are used by classifiers for prediction. However, the quality of different explanation methods has not previously been assessed for detectors of MGT. This study conducts the first systematic evaluation of explanation quality for this task. The dimensions of faithfulness and stability are assessed with five automated experiments, and usefulness is evaluated in a user study. We use a dataset of ChatGPT-generated and human-written documents, and pair predictions of three existing language-model-based detectors with the corresponding SHAP, LIME, and Anchor explanations. We find that SHAP performs best in terms of faithfulness, stability, and in helping users to predict the detector's behavior. In contrast, LIME, perceived as most useful by users, scores the worst in terms of user performance at predicting the detectors' behavior.

Behavioral Testing: Can Large Language Models Implicitly Resolve Ambiguous Entities?

One of the major aspects contributing to the striking performance of large language models (LLMs) is the vast amount of factual knowledge accumulated during pre-training. Yet, many LLMs suffer from self-inconsistency, which raises doubts about their trustworthiness and reliability. In this paper, we focus on entity type ambiguity and analyze current state-of-the-art LLMs for their proficiency and consistency in applying their factual knowledge when prompted for entities under ambiguity. To do so, we propose an evaluation protocol that disentangles knowing from applying knowledge, and test state-of-the-art LLMs on 49 entities. Our experiments reveal that LLMs perform poorly with ambiguous prompts, achieving only 80% accuracy. Our results further demonstrate systematic discrepancies in LLM behavior and their failure to consistently apply information, indicating that the models can exhibit knowledge without being able to utilize it, significant biases for preferred readings, as well as self inconsistencies. Our study highlights the importance of handling entity ambiguity in future for more trustworthy LLMs

Helpful assistant or fruitful facilitator? Investigating how personas affect language model behavior

One way to personalize and steer generations from large language models (LLM) is to assign a persona: a role that describes how the user expects the LLM to behave (e.g., a helpful assistant, a teacher, a woman). This paper investigates how personas affect diverse aspects of model behavior. We assign to seven LLMs 162 personas from 12 categories spanning variables like gender, sexual orientation, and occupation. We prompt them to answer questions from five datasets covering objective (e.g., questions about math and history) and subjective tasks (e.g., questions about beliefs and values). We also compare persona's generations to two baseline settings: a control persona setting with 30 paraphrases of "a helpful assistant" to control for models' prompt sensitivity, and an empty persona setting where no persona is assigned. We find that for all models and datasets, personas show greater variability than the control setting and that some measures of persona behavior generalize across models.

Analysing zero-shot temporal relation extraction on clinical notes using temporal consistency

This paper presents the first study for temporal relation extraction in a zero-shot setting focusing on biomedical text. We employ two types of prompts and five LLMs (GPT-3.5, Mixtral, Llama 2, Gemma, and PMC-LLaMA) to obtain responses about the temporal relations between two events. Our experiments demonstrate that LLMs struggle in the zero-shot setting performing worse than fine-tuned specialized models in terms of F1 score, showing that this is a challenging task for LLMs. We further contribute a novel comprehensive temporal analysis by calculating consistency scores for each LLM. Our findings reveal that LLMs face challenges in providing responses consistent to the temporal properties of uniqueness and transitivity. Moreover, we study the relation between the temporal consistency of an LLM and its accuracy and whether the latter can be improved by solving temporal inconsistencies. Our analysis shows that even when temporal consistency is achieved, the predictions can remain inaccurate.

Exploring prompts to elicit memorization in masked language model-based named entity recognition

Training data memorization in language models impacts model capability (generalization) and safety (privacy risk). This paper focuses on analyzing prompts' impact on detecting the memorization of 6 masked language model-based named entity recognition models. Specifically, we employ a diverse set of 400 automatically generated prompts, and a pairwise dataset where each pair consists of one person's name from the training set and another name out of the set. A prompt completed with a person's name serves as input for getting the model's confidence in predicting this name. Finally, the prompt performance of detecting model memorization is quantified by the percentage of name pairs for which the model has higher confidence for the name from the training set. We show that the performance of different prompts varies by as much as 16 percentage points on the same model, and prompt engineering further increases the gap. Moreover, our experiments demonstrate that prompt performance is model-dependent but does generalize across different name sets. A comprehensive analysis indicates how prompt performance is influenced by prompt properties, contained tokens, and the model's self-attention weights on the prompt.

Specification Overfitting in Artificial Intelligence

Machine learning (ML) and artificial intelligence (AI) approaches are often criticized for their inherent bias and for their lack of control, accountability, and transparency. Consequently, regulatory bodies struggle with containing this technology's potential negative side effects. High-level requirements such as fairness and robustness need to be formalized into concrete specification metrics, imperfect proxies that capture isolated aspects of the underlying requirements. Given possible trade-offs between different metrics and their vulnerability to over-optimization, integrating specification metrics in system development processes is not trivial. This paper defines specification overfitting, a scenario where systems focus excessively on specified metrics to the detriment of high-level requirements and task performance. We present an extensive literature survey to categorize how researchers propose, measure, and optimize specification metrics in several AI fields (e.g., natural language processing, computer vision, reinforcement learning). Using a keyword-based search on papers from major AI conferences and journals between 2018 and mid-2023, we identify and analyze 74 papers that propose or optimize specification metrics. We find that although most papers implicitly address specification overfitting (e.g., by reporting more than one specification metric), they rarely discuss which role specification metrics should play in system development or explicitly define the scope and assumptions behind metric formulations.

Counterfactual Reasoning with Knowledge Graph Embeddings

Knowledge graph embeddings (KGEs) were originally developed to infer true but missing facts in incomplete knowledge repositories. In this paper, we link knowledge graph completion and counterfactual reasoning via our new task CFKGR. We model the original world state as a knowledge graph, hypothetical scenarios as edges added to the graph, and plausible changes to the graph as inferences from logical rules. We create corresponding benchmark datasets, which contain diverse hypothetical scenarios with plausible changes to the original knowledge graph and facts that should be retained. We develop COULDD, a general method for adapting existing knowledge graph embeddings given a hypothetical premise, and evaluate it on our benchmark. Our results indicate that KGEs learn patterns in the graph without explicit training. We further observe that KGEs adapted with COULDD solidly detect plausible counterfactual changes to the graph that follow these patterns. An evaluation on human-annotated data reveals that KGEs adapted with COULDD are mostly unable to recognize changes to the graph that do not follow learned inference rules. In contrast, ChatGPT mostly outperforms KGEs in detecting plausible changes to the graph but has poor knowledge retention. In summary, CFKGR connects two previously distinct areas, namely KG completion and counterfactual reasoning.

Text-Guided Image Clustering

Image clustering divides a collection of images into meaningful groups, typically interpreted post-hoc via human-given annotations. Those are usually in the form of text, begging the question of using text as an abstraction for image clustering. Current image clustering methods, however, neglect the use of generated textual descriptions. We, therefore, propose Text-Guided Image Clustering, i.e., generating text using image captioning and visual question-answering (VQA) models and subsequently clustering the generated text. Further, we introduce a novel approach to inject task- or domain knowledge for clustering by prompting VQA models. Across eight diverse image clustering datasets, our results show that the obtained text representations often outperform image features. Additionally, we propose a counting-based cluster explainability method. Our evaluations show that the derived keyword-based explanations describe clusters better than the respective cluster accuracy suggests. Overall, this research challenges traditional approaches and paves the way for a paradigm shift in image clustering, using generated text.

Functionality learning through specification instructions

Test suites assess natural language processing models' performance on specific functionalities: cases of interest involving model robustness, fairness, or particular linguistic capabilities. They enable fine-grained evaluations of model aspects that would otherwise go unnoticed in standard evaluation datasets, but they do not address the problem of how to fix the failure cases. Previous work has explored functionality learning by fine-tuning models on suite data. While this improves performance on seen functionalities, it often does not generalize to unseen ones and can harm general performance. This paper analyses a fine-tuning-free approach to functionality learning. For each functionality in a suite, we generate a specification instruction that encodes it. We combine the obtained specification instructions to create specification-augmented prompts, which we feed to language models pre-trained on natural instruction data to generate suite predictions. A core aspect of our analysis is to measure the effect that including a set of specifications has on a held-out set of unseen, qualitatively different specifications. Our experiments across four tasks and models ranging from 80M to 175B parameters show that smaller models struggle to follow specification instructions. However, larger models (> 3B params.) can benefit from specifications and even generalize desirable behaviors across functionalities.