We present $\textbf{Platypus}$, a family of fine-tuned and merged Large Language Models (LLMs) that achieves the strongest performance and currently stands at first place in HuggingFace's Open LLM Leaderboard as of the release date of this work. In this work we describe (1) our curated dataset $\textbf{Open-Platypus}$, that is a subset of other open datasets and which $\textit{we release to the public}$ (2) our process of fine-tuning and merging LoRA modules in order to conserve the strong prior of pretrained LLMs, while bringing specific domain knowledge to the surface (3) our efforts in checking for test data leaks and contamination in the training data, which can inform future research. Specifically, the Platypus family achieves strong performance in quantitative LLM metrics across model sizes, topping the global Open LLM leaderboard while using just a fraction of the fine-tuning data and overall compute that are required for other state-of-the-art fine-tuned LLMs. In particular, a 13B Platypus model can be trained on $\textit{a single}$ A100 GPU using 25k questions in 5 hours. This is a testament of the quality of our Open-Platypus dataset, and opens opportunities for more improvements in the field. Project page: https://platypus-llm.github.io

Large language models (LLMs) such as GPT-4, PaLM, and Llama have significantly propelled the generation of AI-crafted text. With rising concerns about their potential misuse, there is a pressing need for AI-generated-text forensics. Neural authorship attribution is a forensic effort, seeking to trace AI-generated text back to its originating LLM. The LLM landscape can be divided into two primary categories: proprietary and open-source. In this work, we delve into these emerging categories of LLMs, focusing on the nuances of neural authorship attribution. To enrich our understanding, we carry out an empirical analysis of LLM writing signatures, highlighting the contrasts between proprietary and open-source models, and scrutinizing variations within each group. By integrating stylometric features across lexical, syntactic, and structural aspects of language, we explore their potential to yield interpretable results and augment pre-trained language model-based classifiers utilized in neural authorship attribution. Our findings, based on a range of state-of-the-art LLMs, provide empirical insights into neural authorship attribution, paving the way for future investigations aimed at mitigating the threats posed by AI-generated misinformation.

Automatic evaluation of machine translation (MT) is a critical tool driving the rapid iterative development of MT systems. While considerable progress has been made on estimating a single scalar quality score, current metrics lack the informativeness of more detailed schemes that annotate individual errors, such as Multidimensional Quality Metrics (MQM). In this paper, we help fill this gap by proposing AutoMQM, a prompting technique which leverages the reasoning and in-context learning capabilities of large language models (LLMs) and asks them to identify and categorize errors in translations. We start by evaluating recent LLMs, such as PaLM and PaLM-2, through simple score prediction prompting, and we study the impact of labeled data through in-context learning and finetuning. We then evaluate AutoMQM with PaLM-2 models, and we find that it improves performance compared to just prompting for scores (with particularly large gains for larger models) while providing interpretability through error spans that align with human annotations.

Adapters and Low-Rank Adaptation (LoRA) are parameter-efficient fine-tuning techniques designed to make the training of language models more efficient. Previous results demonstrated that these methods can even improve performance on some classification tasks. This paper complements the existing research by investigating how these techniques influence the classification performance and computation costs compared to full fine-tuning when applied to multilingual text classification tasks (genre, framing, and persuasion techniques detection; with different input lengths, number of predicted classes and classification difficulty), some of which have limited training data. In addition, we conduct in-depth analyses of their efficacy across different training scenarios (training on the original multilingual data; on the translations into English; and on a subset of English-only data) and different languages. Our findings provide valuable insights into the applicability of the parameter-efficient fine-tuning techniques, particularly to complex multilingual and multilabel classification tasks.

Prompt-based pre-trained language models (PLMs) paradigm have succeeded substantially in few-shot natural language processing (NLP) tasks. However, prior discrete prompt optimization methods require expert knowledge to design the base prompt set and identify high-quality prompts, which is costly, inefficient, and subjective. Meanwhile, existing continuous prompt optimization methods improve the performance by learning the ideal prompts through the gradient information of PLMs, whose high computational cost, and low readability and generalizability are often concerning. To address the research gap, we propose a Dialogue-comprised Policy-gradient-based Discrete Prompt Optimization ($DP_2O$) method. We first design a multi-round dialogue alignment strategy for readability prompt set generation based on GPT-4. Furthermore, we propose an efficient prompt screening metric to identify high-quality prompts with linear complexity. Finally, we construct a reinforcement learning (RL) framework based on policy gradients to match the prompts to inputs optimally. By training a policy network with only 0.67% of the PLM parameter size on the tasks in the few-shot setting, $DP_2O$ outperforms the state-of-the-art (SOTA) method by 1.52% in accuracy on average on four open-source datasets. Moreover, subsequent experiments also demonstrate that $DP_2O$ has good universality, robustness, and generalization ability.

Large Language Models (LLMs) usually suffer from knowledge cutoff or fallacy issues, which means they are unaware of unseen events or generate text with incorrect facts owing to the outdated/noisy data. To this end, many knowledge editing approaches for LLMs have emerged -- aiming to subtly inject/edit updated knowledge or adjust undesired behavior while minimizing the impact on unrelated inputs. Nevertheless, due to significant differences among various knowledge editing methods and the variations in task setups, there is no standard implementation framework available for the community, which hinders practitioners to apply knowledge editing to applications. To address these issues, we propose EasyEdit, an easy-to-use knowledge editing framework for LLMs. It supports various cutting-edge knowledge editing approaches and can be readily apply to many well-known LLMs such as T5, GPT-J, LlaMA, etc. Empirically, we report the knowledge editing results on LlaMA-2 with EasyEdit, demonstrating that knowledge editing surpasses traditional fine-tuning in terms of reliability and generalization. We have released the source code on GitHub at https://github.com/zjunlp/EasyEdit, along with Google Colab tutorials and comprehensive documentation for beginners to get started. Besides, we present an online system for real-time knowledge editing, and a demo video at http://knowlm.zjukg.cn/easyedit.mp4.

A key challenge in professional fact-checking is its limited scalability in relation to the magnitude of false information. While many Natural Language Processing (NLP) tools have been proposed to enhance fact-checking efficiency and scalability, both academic research and fact-checking organizations report limited adoption of such tooling due to insufficient alignment with fact-checker practices, values, and needs. To address this gap, we investigate a co-design method, Matchmaking for AI, which facilitates fact-checkers, designers, and NLP researchers to collaboratively discover what fact-checker needs should be addressed by technology and how. Our co-design sessions with 22 professional fact-checkers yielded a set of 11 novel design ideas. They assist in information searching, processing, and writing tasks for efficient and personalized fact-checking; help fact-checkers proactively prepare for future misinformation; monitor their potential biases; and support internal organization collaboration. Our work offers implications for human-centered fact-checking research and practice and AI co-design research.

Text evaluation has historically posed significant challenges, often demanding substantial labor and time cost. With the emergence of large language models (LLMs), researchers have explored LLMs' potential as alternatives for human evaluation. While these single-agent-based approaches show promise, experimental results suggest that further advancements are needed to bridge the gap between their current effectiveness and human-level evaluation quality. Recognizing that best practices of human evaluation processes often involve multiple human annotators collaborating in the evaluation, we resort to a multi-agent debate framework, moving beyond single-agent prompting strategies. The multi-agentbased approach enables a group of LLMs to synergize with an array of intelligent counterparts, harnessing their distinct capabilities and expertise to enhance efficiency and effectiveness in handling intricate tasks. In this paper, we construct a multi-agent referee team called ChatEval to autonomously discuss and evaluate the quality of generated responses from different models on open-ended questions and traditional natural language generation (NLG) tasks. We derive insights and lessons from practical scenarios where humans instigate group discussions for brainstorming and propose different communication strategies within ChatEval. Our experiments on two benchmark tasks illustrate that ChatEval delivers superior accuracy and correlation in alignment with human assessment. Furthermore, we find that the diverse role prompts (different personas) are essential in the multi-agent debate process; that is, utilizing the same role description in the prompt can lead to a degradation in performance. Our qualitative analysis also shows that ChatEval transcends mere textual scoring, offering a humanmimicking evaluation process for reliable assessments. Evaluating the quality of text generated by language models or written by humans has long been a challenging endeavor, consistently garnering substantial attention (Celikyilmaz et al., 2020). Traditional methodologies predominantly rely on human annotation of texts (Callison-Burch, 2009), an approach considered overly demanding in terms of time and cost.

Finetuning large language models (LLMs) on instructions leads to vast performance improvements on natural language tasks. We apply instruction tuning using code, leveraging the natural structure of Git commits, which pair code changes with human instructions. We compile CommitPack: 4 terabytes of Git commits across 350 programming languages. We benchmark CommitPack against other natural and synthetic code instructions (xP3x, Self-Instruct, OASST) on the 16B parameter StarCoder model, and achieve state-of-the-art performance among models not trained on OpenAI outputs, on the HumanEval Python benchmark (46.2% pass@1). We further introduce HumanEvalPack, expanding the HumanEval benchmark to a total of 3 coding tasks (Code Repair, Code Explanation, Code Synthesis) across 6 languages (Python, JavaScript, Java, Go, C++, Rust). Our models, OctoCoder and OctoGeeX, achieve the best performance across HumanEvalPack among all permissive models, demonstrating CommitPack's benefits in generalizing to a wider set of languages and natural coding tasks. Code, models and data are freely available at https://github.com/bigcode-project/octopack.

LLM-based systems are increasingly deployed in user-facing products in contexts ranging from health (Graber-Stiehl 2023; Harrer 2023) to education (Kasneci et al. 2023), which translates into real-life impacts on thousands of people. Yet despite the many research articles on LLMs, their very definition remains unclear, and much empirical work is based on ideas that lack a solid theoretical grounding or rigorous empirical evidence. To contribute to more rigorous scientific discussion of LLMs, we propose a more precise definition of what factors are necessary and sufficient for an NLP model to be considered an LLM ( 2), and we examine several common claims about LLM properties and functionality ( 3). We then consider the impact that these claims and assumptions have on NLP research ( 4), and we conclude with proposals for maintaining rigor and diversity in NLP research and practice ( 5). Our work draws heavily on both empirical studies and socio-technical critiques of LLMs. 2. What Counts as an LLM? The term "large language model" is used both within the NLP community (e.g. in the call for papers of the Theme Track of EMNLP 2023) as well as in news articles and in U.S. Senate hearings. Despite its ubiquity, its definition is far from clear. For instance, how many parameters should a neural network have to qualify as an LLM? And do only Transformer-type architectures qualify?