SkillAggregation: Reference-free LLM-Dependent Aggregation

Large Language Models (LLMs) are increasingly used to assess NLP tasks due to their ability to generate human-like judgments. Single LLMs were used initially, however, recent work suggests using multiple LLMs as judges yields improved performance. An important step in exploiting multiple judgements is the combination stage, aggregation. Existing methods in NLP either assign equal weight to all LLM judgments or are designed for specific tasks such as hallucination detection. This work focuses on aggregating predictions from multiple systems where no reference labels are available. A new method called SkillAggregation is proposed, which learns to combine estimates from LLM judges without needing additional data or ground truth. It extends the Crowdlayer aggregation method, developed for image classification, to exploit the judge estimates during inference. The approach is compared to a range of standard aggregation methods on HaluEval-Dialogue, TruthfulQA and Chatbot Arena tasks. SkillAggregation outperforms Crowdlayer on all tasks, and yields the best performance over all approaches on the majority of tasks.

Finetuning LLMs for Comparative Assessment Tasks

Automated assessment in natural language generation is a challenging task. Instruction-tuned large language models (LLMs) have shown promise in reference-free evaluation, particularly through comparative assessment. However, the quadratic computational complexity of pairwise comparisons limits its scalability. To address this, efficient comparative assessment has been explored by applying comparative strategies on zero-shot LLM probabilities. We propose a framework for finetuning LLMs for comparative assessment to align the model's output with the target distribution of comparative probabilities. By training on soft probabilities, our approach improves state-of-the-art performance while maintaining high performance with an efficient subset of comparisons.

ASR Error Correction using Large Language Models

Error correction (EC) models play a crucial role in refining Automatic Speech Recognition (ASR) transcriptions, enhancing the readability and quality of transcriptions. Without requiring access to the underlying code or model weights, EC can improve performance and provide domain adaptation for black-box ASR systems. This work investigates the use of large language models (LLMs) for error correction across diverse scenarios. 1-best ASR hypotheses are commonly used as the input to EC models. We propose building high-performance EC models using ASR N-best lists which should provide more contextual information for the correction process. Additionally, the generation process of a standard EC model is unrestricted in the sense that any output sequence can be generated. For some scenarios, such as unseen domains, this flexibility may impact performance. To address this, we introduce a constrained decoding approach based on the N-best list or an ASR lattice. Finally, most EC models are trained for a specific ASR system requiring retraining whenever the underlying ASR system is changed. This paper explores the ability of EC models to operate on the output of different ASR systems. This concept is further extended to zero-shot error correction using LLMs, such as ChatGPT. Experiments on three standard datasets demonstrate the efficacy of our proposed methods for both Transducer and attention-based encoder-decoder ASR systems. In addition, the proposed method can serve as an effective method for model ensembling.

Controlling Whisper: Universal Acoustic Adversarial Attacks to Control Speech Foundation Models

Speech enabled foundation models, either in the form of flexible speech recognition based systems or audio-prompted large language models (LLMs), are becoming increasingly popular. One of the interesting aspects of these models is their ability to perform tasks other than automatic speech recognition (ASR) using an appropriate prompt. For example, the OpenAI Whisper model can perform both speech transcription and speech translation. With the development of audio-prompted LLMs there is the potential for even greater control options. In this work we demonstrate that with this greater flexibility the systems can be susceptible to model-control adversarial attacks. Without any access to the model prompt it is possible to modify the behaviour of the system by appropriately changing the audio input. To illustrate this risk, we demonstrate that it is possible to prepend a short universal adversarial acoustic segment to any input speech signal to override the prompt setting of an ASR foundation model. Specifically, we successfully use a universal adversarial acoustic segment to control Whisper to always perform speech translation, despite being set to perform speech transcription. Overall, this work demonstrates a new form of adversarial attack on multi-tasking speech enabled foundation models that needs to be considered prior to the deployment of this form of model.

Cross-Lingual Transfer Learning for Speech Translation

There has been increasing interest in building multilingual foundation models for NLP and speech research. Zero-shot cross-lingual transfer has been demonstrated on a range of NLP tasks where a model fine-tuned on task-specific data in one language yields performance gains in other languages. Here, we explore whether speech-based models exhibit the same transfer capability. Using Whisper as an example of a multilingual speech foundation model, we examine the utterance representation generated by the speech encoder. Despite some language-sensitive information being preserved in the audio embedding, words from different languages are mapped to a similar semantic space, as evidenced by a high recall rate in a speech-to-speech retrieval task. Leveraging this shared embedding space, zero-shot cross-lingual transfer is demonstrated in speech translation. When the Whisper model is fine-tuned solely on English-to-Chinese translation data, performance improvements are observed for input utterances in other languages. Additionally, experiments on low-resource languages show that Whisper can perform speech translation for utterances from languages unseen during pre-training by utilizing cross-lingual representations.

CrossCheckGPT: Universal Hallucination Ranking for Multimodal Foundation Models

Multimodal foundation models are prone to hallucination, generating outputs that either contradict the input or are not grounded by factual information. Given the diversity in architectures, training data and instruction tuning techniques, there can be large variations in systems' susceptibility to hallucinations. To assess system hallucination robustness, hallucination ranking approaches have been developed for specific tasks such as image captioning, question answering, summarization, or biography generation. However, these approaches typically compare model outputs to gold-standard references or labels, limiting hallucination benchmarking for new domains. This work proposes "CrossCheckGPT", a reference-free universal hallucination ranking for multimodal foundation models. The core idea of CrossCheckGPT is that the same hallucinated content is unlikely to be generated by different independent systems, hence cross-system consistency can provide meaningful and accurate hallucination assessment scores. CrossCheckGPT can be applied to any model or task, provided that the information consistency between outputs can be measured through an appropriate distance metric. Focusing on multimodal large language models that generate text, we explore two information consistency measures: CrossCheck-explicit and CrossCheck-implicit. We showcase the applicability of our method for hallucination ranking across various modalities, namely the text, image, and audio-visual domains. Further, we propose the first audio-visual hallucination benchmark, "AVHalluBench", and illustrate the effectiveness of CrossCheckGPT, achieving correlations of 98% and 89% with human judgements on MHaluBench and AVHalluBench, respectively.

Question-Based Retrieval using Atomic Units for Enterprise RAG

Enterprise retrieval augmented generation (RAG) offers a highly flexible framework for combining powerful large language models (LLMs) with internal, possibly temporally changing, documents. In RAG, documents are first chunked. Relevant chunks are then retrieved for a specific user query, which are passed as context to a synthesizer LLM to generate the query response. However, the retrieval step can limit performance, as incorrect chunks can lead the synthesizer LLM to generate a false response. This work proposes a zero-shot adaptation of standard dense retrieval steps for more accurate chunk recall. Specifically, a chunk is first decomposed into atomic statements. A set of synthetic questions are then generated on these atoms (with the chunk as the context). Dense retrieval involves finding the closest set of synthetic questions, and associated chunks, to the user query. It is found that retrieval with the atoms leads to higher recall than retrieval with chunks. Further performance gain is observed with retrieval using the synthetic questions generated over the atoms. Higher recall at the retrieval step enables higher performance of the enterprise LLM using the RAG pipeline.

Efficient LLM Comparative Assessment: a Product of Experts Framework for Pairwise Comparisons

LLM-as-a-judge approaches are a practical and effective way of assessing a range of text tasks, aligning with human judgements especially when applied in a comparative assessment fashion. However, when using pairwise comparisons to rank a set of candidates the computational costs scale quadratically with the number of candidates, which can have practical limitations. This paper introduces a Product of Expert (PoE) framework for efficient LLM Comparative Assessment. Here individual comparisons are considered experts that provide information on a pair's score difference. The PoE framework combines the information from these experts to yield an expression that can be maximized with respect to the underlying set of candidates, and is highly flexible where any form of expert can be assumed. When Gaussian experts are used one can derive simple closed-form solutions for the optimal candidate ranking, as well as expressions for selecting which comparisons should be made to maximize the probability of this ranking. Our approach enables efficient comparative assessment, where by using only a small subset of the possible comparisons, one can generate score predictions that correlate as well to human judgements as the predictions when all comparisons are used. We evaluate the approach on multiple NLG tasks and demonstrate that our framework can yield considerable computational savings when performing pairwise comparative assessment. When N is large, with as few as 2% of comparisons the PoE solution can achieve similar performance to when all comparisons are used.

Muting Whisper: A Universal Acoustic Adversarial Attack on Speech Foundation Models

Recent developments in large speech foundation models like Whisper have led to their widespread use in many automatic speech recognition (ASR) applications. These systems incorporate `special tokens' in their vocabulary, such as $\texttt{<endoftext>}$, to guide their language generation process. However, we demonstrate that these tokens can be exploited by adversarial attacks to manipulate the model's behavior. We propose a simple yet effective method to learn a universal acoustic realization of Whisper's $\texttt{<endoftext>}$ token, which, when prepended to any speech signal, encourages the model to ignore the speech and only transcribe the special token, effectively `muting' the model. Our experiments demonstrate that the same, universal 0.64-second adversarial audio segment can successfully mute a target Whisper ASR model for over 97\% of speech samples. Moreover, we find that this universal adversarial audio segment often transfers to new datasets and tasks. Overall this work demonstrates the vulnerability of Whisper models to `muting' adversarial attacks, where such attacks can pose both risks and potential benefits in real-world settings: for example the attack can be used to bypass speech moderation systems, or conversely the attack can also be used to protect private speech data.

Question Difficulty Ranking for Multiple-Choice Reading Comprehension

Multiple-choice (MC) tests are an efficient method to assess English learners. It is useful for test creators to rank candidate MC questions by difficulty during exam curation. Typically, the difficulty is determined by having human test takers trial the questions in a pretesting stage. However, this is expensive and not scalable. Therefore, we explore automated approaches to rank MC questions by difficulty. However, there is limited data for explicit training of a system for difficulty scores. Hence, we compare task transfer and zero-shot approaches: task transfer adapts level classification and reading comprehension systems for difficulty ranking while zero-shot prompting of instruction finetuned language models contrasts absolute assessment against comparative. It is found that level classification transfers better than reading comprehension. Additionally, zero-shot comparative assessment is more effective at difficulty ranking than the absolute assessment and even the task transfer approaches at question difficulty ranking with a Spearman's correlation of 40.4%. Combining the systems is observed to further boost the correlation.