Malta

Large Language Models (LLMs) have taken the research field of Natural Language Processing by storm. Researchers are not only investigating their capabilities and possible applications, but also their weaknesses and how they may be exploited. This has resulted in various attacks and &quot;jailbreaking&quot; approaches that have gained large interest within the community. The vulnerability of LLMs to certain types of input may pose major risks regarding the real-world usage of LLMs in productive operations. We therefore investigate the relationship between a LLM&#39;s uncertainty and its vulnerability to jailbreaking attacks. To this end, we focus on a probabilistic point of view of uncertainty and employ a state-of-the art open-source LLM. We investigate an attack that is based on linguistic obfuscation. Our results indicate that the model is subject to a higher level of uncertainty when confronted with manipulated prompts that aim to evade security mechanisms. This study lays the foundation for future research into the link between model uncertainty and its vulnerability to jailbreaks.

EACL 2024

Linguistic Obfuscation Attacks and Large Language Model Uncertainty

jailbreak

uncertainty

attack

Large Language Models (LLMs) have taken the research field of Natural Language Processing by storm. Researchers are not only investigating their capabilities and possible applications, but also their weaknesses and how they may be exploited. This has resulted in various attacks and "jailbreaking" approaches that have gained large interest within the community. The vulnerability of LLMs to certain types of input may pose major risks regarding the real-world usage of LLMs in productive operations. We therefore investigate the relationship between a LLM's uncertainty and its vulnerability to jailbreaking attacks. To this end, we focus on a probabilistic point of view of uncertainty and employ a state-of-the art open-source LLM. We investigate an attack that is based on linguistic obfuscation. Our results indicate that the model is subject to a higher level of uncertainty when confronted with manipulated prompts that aim to evade security mechanisms. This study lays the foundation for future research into the link between model uncertainty and its vulnerability to jailbreaks.

workshop paper

Welcome to the 18th Conference of the European Chapter of the Association for Computational Linguistics (EACL). Continuing its mission of expanding and involving the science community of all European countries, EACL has selected the Malta community for the 18th EACL. Considering the importance of physical interaction among researchers, the conference will be held at the Hotel Radisson Blu, St. Julians, in Malta, from 17 to 22 of March, 2024. The conference will also feature the possibility to participate virtually. As the flagship European conference in the field of computational linguistics, EACL welcomes European and international researchers covering a broad spectrum of research areas that are concerned with computational approaches to natural language. 
#### [Conference Handbook](https://drive.google.com/file/d/13Wn38q0yev6U3RTgnf4Tn5T9GuAuZ-xW/view?usp=sharing) *downloadable*

To access the EACL 2024 event page you need to register [**here**](https://acl.swoogo.com/EACL2024). 

Please register!

As the flagship European conference in the field of computational linguistics, EACL welcomes European and international researchers covering a broad spectrum of research areas that are concerned with computational approaches to natural language.

Automatically generated summaries can be evaluated along different dimensions, one being how faithfully the uncertainty from the source text is conveyed in the summary. We present a study on uncertainty alignment in automatic summarization, starting from a two-tier lexical and semantic categorization of linguistic expression of uncertainty, which we used to annotate source texts and automatically generate summaries. We collected a diverse dataset including news articles and personal blogs and generated summaries using GPT-4. Source texts and summaries were annotated based on our two-tier taxonomy using a markup language. The automatic annotation was refined and validated by subsequent iterations based on expert input. We propose a method to evaluate the fidelity of uncertainty transfer in text summarization. The method capitalizes on a small amount of expert annotations and on the capabilities of Large language models (LLMs) to evaluate how the uncertainty of the source text aligns with the uncertainty expressions in the summary.

Aligning Uncertainty: Leveraging LLMs to Analyze Uncertainty Transfer in Text Summarization

Sequence labeling is a core task in text understanding for IE/IR systems. Text generation models have increasingly become the go-to solution for such tasks (e.g., entity extraction and dialog slot filling). While most research has focused on the labeling accuracy, a key aspect -- of vital practical importance -- has slipped through the cracks: understanding model confidence. More specifically, we lack a principled understanding of how to reliably gauge the confidence of a model in its predictions for each labeled span. This paper aims to provide some empirical insights on estimating model confidence for generative sequence labeling. Most notably, we find that simply using the decoder's output probabilities \textbf{is not} the best in realizing well-calibrated confidence estimates. As verified over six public datasets of different tasks, we show that our proposed approach -- which leverages statistics from top-$k$ predictions by a beam search -- significantly reduces calibration errors of the predictions of a generative sequence labeling model.

How Does Beam Search improve Span-Level Confidence Estimation in Generative Sequence Labeling?

Learning from human feedback can improve models for text generation or passage ranking, aligning them better to a user's needs. Data is often collected by asking users to compare alternative outputs to a given input, which may require a large number of comparisons to learn a ranking function. The amount of comparisons needed can be reduced using Bayesian Optimisation (BO) to query the user about only the most promising candidate outputs. Previous applications of BO to text ranking relied on shallow surrogate models to learn ranking functions over candidate outputs, and were therefore unable to fine-tune rankers based on deep, pretrained language models. This paper leverages Bayesian deep learning (BDL) to adapt pretrained language models to highly specialised text ranking tasks, using BO to tune the model with a small number of pairwise preferences between candidate outputs. We apply our approach to community question answering (cQA) and extractive multi-document summarisation (MDS) with simulated noisy users, finding that our BDL approach significantly outperforms both a shallow Gaussian process model and traditional active learning with a standard deep neural network, while remaining robust to noise in the user feedback.

Efficiently Acquiring Human Feedback with Bayesian Deep Learning

The highest probability sequences of most neural language generation models tend to be degenerate in some way, a problem known as the inadequacy of the mode. While many approaches to tackling particular aspects of the problem exist, such as dealing with too short sequences or excessive repetitions, explanations of why it occurs in the first place are rarer and do not agree with each other. We believe none of the existing explanations paint a complete picture. In this position paper, we want to bring light to the incredible complexity of the modelling task and the problems that generalising to previously unseen contexts bring. We argue that our desire for models to generalise to contexts it has never observed before is exactly what leads to spread of probability mass and inadequate modes. While we do not claim that adequate modes are impossible, we argue that they are not to be expected either.

The Effect of Generalisation on the Inadequacy of the Mode

Researchers have raised awareness about the harms of aggregating labels especially in subjective tasks that naturally contain disagreements among human annotators. In this work we show that models that are only provided aggregated labels show low confidence on high-disagreement data instances. While previous studies consider such instances as mislabeled, we argue that the reason the high-disagreement text instances have been hard-to-learn is that the conventional aggregated models underperform in extracting useful signals from subjective tasks. Inspired by recent studies demonstrating the effectiveness of learning from raw annotations, we investigate classifying using Multiple Ground Truth (Multi-GT) approaches. Our experiments show an improvement of confidence for the high-disagreement instances.

Don't Blame the Data, Blame the Model: Understanding Noise and Bias When Learning from Subjective Annotations

Large Language Models have emerged as prime candidates to tackle misinformation mitigation. However, existing approaches struggle with hallucinations and overconfident predictions. We propose an uncertainty quantification framework that leverages both direct confidence elicitation and sampled-based consistency methods to provide better calibration for NLP misinformation mitigation solutions. We first investigate the calibration of sample-based consistency methods that exploit distinct features of consistency across sample sizes and stochastic levels. Next, we evaluate the performance and distributional shift of a robust numeric verbalization prompt across single vs. two-step confidence elicitation procedure. We also compare the performance of the same prompt with different versions of GPT and different numerical scales. Finally, we combine the sample-based consistency and verbalized methods to propose a hybrid framework that yields a better uncertainty estimation for GPT models. Overall, our work proposes novel uncertainty quantification methods that will improve the reliability of Large Language Models in misinformation mitigation applications.

Combining Confidence Elicitation and Sample-based Methods for Uncertainty Quantification in Misinformation Mitigation

This paper addresses the unique challenges associated with uncertainty quantification in AI models when applied to patient-facing contexts within healthcare. Unlike traditional eXplainable Artificial Intelligence (XAI) methods tailored for model developers or domain experts, additional considerations of communicating in natural language, its presentation and evaluating understandability are necessary. We identify the challenges in communication model performance, confidence, reasoning and unknown knowns using natural language in the context of risk prediction. We propose a design aimed at addressing these challenges, focusing on the specific application of in-vitro fertilisation outcome prediction.

Linguistically Communicating Uncertainty in Patient-Facing Risk Prediction Models

This paper introduces a novel decision-making framework that promotes consistency among decisions made by diverse models while utilizing external knowledge. Leveraging the Integer Linear Programming(ILP) framework, we map predictions from various models into globally normalized and comparable values by incorporating information about decisions' prior probability, confidence(uncertainty), and the models' expected accuracy. Our empirical study demonstrates the superiority of our approach over conventional baselines on multiple datasets.

Consistent Joint Decision-Making with Heterogeneous Learning Models

In this paper, we evaluate the ability of transformer-based language models in reasoning over uncertain text that includes uncertain rules of reasoning. We cover pre-trained language models (PLMs) and the newer large language models (LLMs). Our evaluation results show that both generations of language models struggle with reasoning over uncertain text. We focus on PLMs and propose a novel Neuro-Symbolic fine-tuning approach, Probabilistic Constraint Training (PCT), incorporating probabilistic logical rules as constraints during fine-tuning. To assess the effectiveness of PCT, we utilize the related corpora and, additionally, create a new and more challenging benchmark that, unlike the previous ones, uses instance-specific rules. Our study demonstrates the potential of PCT, the pioneer method that improves the transformer-based language model's accuracy and explainability of the probabilistic logical reasoning process. Furthermore, PCT equips these models to effectively handle novel situations, including higher reasoning depth, new domains, and complex probabilistic structures.

Teaching Probabilistic Logical Reasoning to Transformers

In this paper, we evaluate two different natural language processing (NLP) approaches to solve a paradigmatic task for computational literary studies (CLS): the recognition of knowledge transfer in literary texts. We focus on the question of how adequately do large language models capture the transfer of knowledge about family relations in German drama texts when this transfer is treated as a classification or textual entailment task using in-context learning (ICL). We find that a 13 billion parameter Llama 2 model performs best on the first task, while GPT-4 performs best on the latter task. However, all models achieve relatively low scores compared to standard NLP benchmark results, struggle from inconsistencies with small changes in prompts and are often not able to make simple inferences beyond the textual surface, which is why an unreflected generic use of ICL in the CLS seems still not advisable.

Evaluating In-Context Learning for Computational Literary Studies: A Case Study Based on the Automatic Recognition of Knowledge Transfer in German Drama

Premium content

Linguistic Obfuscation Attacks and Large Language Model Uncertainty

Downloads

Next from EACL 2024

Aligning Uncertainty: Leveraging LLMs to Analyze Uncertainty Transfer in Text Summarization

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES