Retrieval augmented models show promise in enhancing traditional language models by improving their contextual understanding, integrating private data, and reducing hallucination. However, the processing time required for retrieval augmented large language models poses a challenge when applying them to tasks that require real-time responses, such as composition assistance. To overcome this limitation, we propose the Hybrid Retrieval-Augmented Generation (HybridRAG) framework that leverages a hybrid setting that combines both client and cloud models. HybridRAG incorporates retrieval-augmented memory generated asynchronously by a Large Language Model (LLM) in the cloud. By integrating this retrieval augmented memory, the client model acquires the capability to generate highly effective responses, benefiting from the LLM's capabilities. Furthermore, through asynchronous memory integration, the client model is capable of delivering real-time responses to user requests without the need to wait for memory synchronization from the cloud. Our experiments on Wikitext and Pile subsets show that HybridRAG achieves lower latency than a cloud-based retrieval-augmented LLM, while outperforming client-only models in utility.

Simile detection is a valuable task for many natural language processing (NLP)-based applications, particularly in the field of literature. However, existing research on simile detection often relies on corpora that are limited in size and do not adequately represent the full range of simile forms. To address this issue, we propose a simile data augmentation method based on \textbf{W}ord replacement And Sentence completion using the GPT-2 language model. Our iterative process called I-WAS, is designed to improve the quality of the augmented sentences. To better evaluate the performance of our method in real-world applications, we have compiled a corpus containing a more diverse set of simile forms for experimentation. Our experimental results demonstrate the effectiveness of our proposed data augmentation method for simile detection.

The analysis of public affairs documents is crucial for citizens as it promotes transparency, accountability, and informed decision-making. It allows citizens to understand government policies, participate in public discourse, and hold representatives accountable. This is crucial, and sometimes a matter of life or death, for companies whose operation depend on certain regulations. Large Language Models (LLMs) have the potential to greatly enhance the analysis of public affairs documents by effectively processing and understanding the complex language used in such documents. In this work, we analyze the performance of LLMs in classifying public affairs documents. As a natural multi-label task, the classification of these documents presents important challenges. In this work, we use a regex-powered tool to collect a database of public affairs documents with more than 33K samples and 22.5M tokens. Our experiments assess the performance of 4 different Spanish LLMs to classify up to 30 different topics in the data in different configurations. The results shows that LLMs can be of great use to process domain-specific documents, such as those in the domain of public affairs.

Large Language Models (LLMs) have shown outstanding performance across wide range of downstream tasks. This competency is attributed to their substantial parameter size and pre-training on extensive corpus. Moreover, LLMs have exhibited enhanced reasoning capabilities in tackling complex reasoning tasks, owing to the utilization of a method named ``Chain-of-Thought (CoT) prompting''. This method is designed to generate intermediate reasoning steps that guide the inference of the final answer. However, it is essential to highlight that these advanced reasoning abilities appear to emerge in models with a minimum of 10 billion parameters, thereby limiting its efficacy in situations where computational resources are constrained. In this paper, we investigate the possibility of transferring the reasoning capabilities of LLMs to smaller models via knowledge distillation. Specifically, we propose Sci-CoT, a two-stage framework that separates the processes of generating rationales and inferring answers. This method enables a more efficient use of rationales during the answer inference stage, leading to improved performance on scientific question-answering tasks. Utilizing Sci-CoT, our 80-million parameter model is able to exceed the performance of BLOOM-176B in the ARC-Easy dataset under the few shot setting.

Autonomous conversational agents, i.e. chatbots, are becoming an increasingly common mechanism for enterprises to provide support to customers and partners. In order to rate chatbots, especially ones powered by Generative AI tools like Large Language Models (LLMs) we need to be able to accurately assess their performance. This is where chatbot benchmarking becomes important. In this paper, we propose the use of a novel benchmark that we call the E2E (End to End) benchmark, and show how the E2E benchmark can be used to evaluate accuracy and usefulness of the answers provided by chatbots, especially ones powered by LLMs. We evaluate an example chatbot at different levels of sophistication based on both our E2E benchmark, as well as other available metrics commonly used in the state of art, and observe that the proposed benchmark show better results compared to others. In addition, while some metrics proved to be unpredictable, the metric associated with the E2E benchmark, which uses cosine similarity performed well in evaluating chatbots. The performance of our best models shows that there are several benefits of using the cosine similarity score as a metric in the E2E benchmark.

Recent advances with large language models (LLM) illustrate their diverse capabilities. We propose a novel algorithm, staged speculative decoding, to accelerate LLM inference in small-batch, on-device scenarios. We address the low arithmetic intensity of small-batch inference by improving upon previous work in speculative decoding. First, we restructure the speculative batch as a tree, which reduces generation costs and increases the expected tokens per batch. Second, we add a second stage of speculative decoding. Taken together, we reduce single-batch decoding latency by 3.16x with a 762M parameter GPT-2-L model while perfectly preserving output quality.

In recent years, federated learning (FL) has become a very popular paradigm for training distributed, large-scale, and privacy-preserving machine learning (ML) systems. In contrast to standard ML, where data must be collected at the exact location where training is performed, FL takes advantage of the computational capabilities of millions of edge devices to collaboratively train a shared, global model without disclosing their local private data. Specifically, in a typical FL system, the central server acts only as an orchestrator; it iteratively gathers and aggregates all the local models trained by each client on its private data until convergence. Although FL undoubtedly has several benefits over traditional ML (e.g., it protects private data ownership by design), it suffers from several weaknesses. One of the most critical challenges is to overcome the centralized orchestration of the classical FL client-server architecture, which is known to be vulnerable to single-point-of-failure risks and man-in-the-middle attacks, among others. To mitigate such exposure, decentralized FL solutions have emerged where all FL clients cooperate and communicate without a central server. This survey comprehensively summarizes and reviews existing decentralized FL approaches proposed in the literature. Furthermore, it identifies emerging challenges and suggests promising research directions in this under-explored domain.

As large language models improve, there is increasing interest in techniques that leverage these models' capabilities to refine their own outputs. In this work, we introduce Shepherd, a language model specifically tuned to critique responses and suggest refinements, extending beyond the capabilities of an untuned model to identify diverse errors and provide suggestions to remedy them. At the core of our approach is a high quality feedback dataset, which we curate from community feedback and human annotations. Even though Shepherd is small (7B parameters), its critiques are either equivalent or preferred to those from established models including ChatGPT. Using GPT-4 for evaluation, Shepherd reaches an average win-rate of 53-87% compared to competitive alternatives. In human evaluation, Shepherd strictly outperforms other models and on average closely ties with ChatGPT.

Recently, large language models (LLMs) fine-tuned to follow human instruction have exhibited significant capabilities in various English NLP tasks. However, their performance in grammatical error correction (GEC) tasks, particularly in non-English languages, remains significantly unexplored. In this paper, we delve into abilities of instruction fine-tuned LLMs in Arabic GEC, a task made complex due to Arabic's rich morphology. Our findings suggest that various prompting methods, coupled with (in-context) few-shot learning, demonstrate considerable effectiveness, with GPT-4 achieving up to $65.49$ F\textsubscript{1} score under expert prompting (approximately $5$ points higher than our established baseline). This highlights the potential of LLMs in low-resource settings, offering a viable approach for generating useful synthetic data for model training. Despite these positive results, we find that instruction fine-tuned models, regardless of their size, significantly underperform compared to fully fine-tuned models of significantly smaller sizes. This disparity highlights a substantial room for improvements for LLMs. Inspired by methods from low-resource machine translation, we also develop a method exploiting synthetic data that significantly outperforms previous models on two standard Arabic benchmarks. Our work sets new SoTA for Arabic GEC, with $72.19\%$ and $73.26$ F$_{1}$ on the 2014 and 2015 QALB datasets, respectively.

This paper presents AutoHint, a novel framework for automatic prompt engineering and optimization for Large Language Models (LLM). While LLMs have demonstrated remarkable ability in achieving high-quality annotation in various tasks, the key to applying this ability to specific tasks lies in developing high-quality prompts. Thus we propose a framework to inherit the merits of both in-context learning and zero-shot learning by incorporating enriched instructions derived from input-output demonstrations to optimize original prompt. We refer to the enrichment as the hint and propose a framework to automatically generate the hint from labeled data. More concretely, starting from an initial prompt, our method first instructs a LLM to deduce new hints for selected samples from incorrect predictions, and then summarizes from per-sample hints and adds the results back to the initial prompt to form a new, enriched instruction. The proposed method is evaluated on the BIG-Bench Instruction Induction dataset for both zero-shot and few-short prompts, where experiments demonstrate our method is able to significantly boost accuracy for multiple tasks.