Fuzzing has achieved tremendous success in discovering bugs and vulnerabilities in various software systems. Systems under test (SUTs) that take in programming or formal language as inputs, e.g., compilers, runtime engines, constraint solvers, and software libraries with accessible APIs, are especially important as they are fundamental building blocks of software development. However, existing fuzzers for such systems often target a specific language, and thus cannot be easily applied to other languages or even other versions of the same language. Moreover, the inputs generated by existing fuzzers are often limited to specific features of the input language, and thus can hardly reveal bugs related to other or new features. This paper presents Fuzz4All, the first fuzzer that is universal in the sense that it can target many different input languages and many different features of these languages. The key idea behind Fuzz4All is to leverage large language models (LLMs) as an input generation and mutation engine, which enables the approach to produce diverse and realistic inputs for any practically relevant language. To realize this potential, we present a novel autoprompting technique, which creates LLM prompts that are wellsuited for fuzzing, and a novel LLM-powered fuzzing loop, which iteratively updates the prompt to create new fuzzing inputs. We evaluate Fuzz4All on nine systems under test that take in six different languages (C, C++, Go, SMT2, Java and Python) as inputs. The evaluation shows, across all six languages, that universal fuzzing achieves higher coverage than existing, language-specific fuzzers. Furthermore, Fuzz4All has identified 76 bugs in widely used systems, such as GCC, Clang, Z3, CVC5, OpenJDK, and the Qiskit quantum computing platform, with 47 bugs already confirmed by developers as previously unknown.

In recent years, there have been significant breakthroughs in the field of natural language processing, particularly with the development of large language models (LLMs). These LLMs have showcased remarkable capabilities on various benchmarks. In the healthcare field, the exact role LLMs and other future AI models will play remains unclear. There is a potential for these models in the future to be used as part of adaptive physician training, medical co-pilot applications, and digital patient interaction scenarios. The ability of AI models to participate in medical training and patient care will depend in part on their mastery of the knowledge content of specific medical fields. This study investigated the medical knowledge capability of LLMs, specifically in the context of internal medicine subspecialty multiple-choice test-taking ability. We compared the performance of several open-source LLMs (Koala 7B, Falcon 7B, Stable-Vicuna 13B, and Orca Mini 13B), to GPT-4 and Claude 2 on multiple-choice questions in the field of Nephrology. Nephrology was chosen as an example of a particularly conceptually complex subspecialty field within internal medicine. The study was conducted to evaluate the ability of LLM models to provide correct answers to nephSAP (Nephrology Self-Assessment Program) multiple-choice questions. The overall success of open-sourced LLMs in answering the 858 nephSAP multiple-choice questions correctly was 17.1% - 25.5%. In contrast, Claude 2 answered 54.4% of the questions correctly, whereas GPT-4 achieved a score of 73.3%. We show that current widely used open-sourced LLMs do poorly in their ability for zero-shot reasoning when compared to GPT-4 and Claude 2. The findings of this study potentially have significant implications for the future of subspecialty medical training and patient care.

Researchers, practitioners, engineers and students can find themselves grappling with a multitude of acronyms and intricate terminology with information spread across a large number of documents, which can prove to be an onerous and time-consuming task to work with and develop standards-compliant systems. For example, an engineering team working on implementing registration request procedure as a part of building 5G virtual core would need to identify all the relevant technical specifications from among thousands of such documents, and understand the call flow and message formats as described in those specifications. Table 1 provides several examples of such user stories. The current method of acquiring this information involves sifting through numerous webpages and technical specification documents. While this approach provides extensive comprehension of a topic from various sources, it can also be very time-intensive and tedious to identify multiple relevant sources, gather information from them and synthesize it [22]. The emergence of foundation models [6] like ChatGPT [35] presents a promising prospect for solving this problem as they represent a significant advancement in providing synthesized, readily comprehensible answers to user queries related to wireless communication specifications and technologies. However, despite the usefulness of state-of-the-art foundation large language models (LLMs) in answering many queries related to modern wireless communication technologies, they offer irrelevant or inaccurate responses to many of these queries. For example, as shown in Figure 1(a), when prompted about'what is numerology in 5G', ChatGPT (Feb 2023) describes that numerology is related to mystical significance of numbers and has no connection to 5G. Similarly, when prompted about'the number of unique values physical identity can take in 5G', it responds that'PCI consists of a 3-bit value ranging from 0 to 503', which is inaccurate and also non-sensible as 3-bit value cannot take 504 different values.

Robots are becoming an essential part of many operations including marine exploration or environmental monitoring. However, the underwater environment presents many challenges, including high pressure, limited visibility, and harsh conditions that can damage equipment. Real-world experimentation can be expensive and difficult to execute. Therefore, it is essential to simulate the performance of underwater robots in comparable environments to ensure their optimal functionality within practical real-world contexts.OysterSim generates photo-realistic images and segmentation masks of objects in marine environments, providing valuable training data for underwater computer vision applications. By integrating ChatGPT into underwater simulations, users can convey their thoughts effortlessly and intuitively create desired underwater environments without intricate coding. \invis{Moreover, researchers can realize substantial time and cost savings by evaluating their algorithms across diverse underwater conditions in the simulation.} The objective of ChatSim is to integrate Large Language Models (LLM) with a simulation environment~(OysterSim), enabling direct control of the simulated environment via natural language input. This advancement can greatly enhance the capabilities of underwater simulation, with far-reaching benefits for marine exploration and broader scientific research endeavors.

N-gram matching-based evaluation metrics, such as BLEU and chrF, are widely utilized across a range of natural language generation (NLG) tasks. However, recent studies have revealed a weak correlation between these matching-based metrics and human evaluations, especially when compared with neural-based metrics like BLEURT. In this paper, we conjecture that the performance bottleneck in matching-based metrics may be caused by the limited diversity of references. To address this issue, we propose to utilize \textit{multiple references} to enhance the consistency between these metrics and human evaluations. Within the WMT Metrics benchmarks, we observe that the multi-references F200spBLEU surpasses the conventional single-reference one by an accuracy improvement of 7.2\%. Remarkably, it also exceeds the neural-based BERTscore by an accuracy enhancement of 3.9\%. Moreover, we observe that the data leakage issue in large language models (LLMs) can be mitigated to a large extent by our multi-reference metric. We release the code and data at \url{https://github.com/SefaZeng/LLM-Ref}

Cross-domain and cross-compositional generalization of Text-to-SQL semantic parsing is a challenging task. Existing Large Language Model (LLM) based solutions rely on inference-time retrieval of few-shot exemplars from the training set to synthesize a run-time prompt for each Natural Language (NL) test query. In contrast, we devise an algorithm which performs offline sampling of a minimal set-of few-shots from the training data, with complete coverage of SQL clauses, operators and functions, and maximal domain coverage within the allowed token length. This allows for synthesis of a fixed Generic Prompt (GP), with a diverse set-of exemplars common across NL test queries, avoiding expensive test time exemplar retrieval. We further auto-adapt the GP to the target database domain (DA-GP), to better handle cross-domain generalization; followed by a decomposed Least-To-Most-Prompting (LTMP-DA-GP) to handle cross-compositional generalization. The synthesis of LTMP-DA-GP is an offline task, to be performed one-time per new database with minimal human intervention. Our approach demonstrates superior performance on the KaggleDBQA dataset, designed to evaluate generalizability for the Text-to-SQL task. We further showcase consistent performance improvement of LTMP-DA-GP over GP, across LLMs and databases of KaggleDBQA, highlighting the efficacy and model agnostic benefits of our prompt based adapt and decompose approach.

In this work, we propose Retentive Network (RetNet) as a foundation architecture for large language models, simultaneously achieving training parallelism, low-cost inference, and good performance. We theoretically derive the connection between recurrence and attention. Then we propose the retention mechanism for sequence modeling, which supports three computation paradigms, i.e., parallel, recurrent, and chunkwise recurrent. Specifically, the parallel representation allows for training parallelism. The recurrent representation enables low-cost $O(1)$ inference, which improves decoding throughput, latency, and GPU memory without sacrificing performance. The chunkwise recurrent representation facilitates efficient long-sequence modeling with linear complexity, where each chunk is encoded parallelly while recurrently summarizing the chunks. Experimental results on language modeling show that RetNet achieves favorable scaling results, parallel training, low-cost deployment, and efficient inference. The intriguing properties make RetNet a strong successor to Transformer for large language models. Code will be available at https://aka.ms/retnet.

Informal reasoning ability is the ability to reason based on common sense, experience, and intuition.Humans use informal reasoning every day to extract the most influential elements for their decision-making from a large amount of life-like information.With the rapid development of language models, the realization of general artificial intelligence has emerged with hope. Given the outstanding informal reasoning ability of humans, how much informal reasoning ability language models have has not been well studied by scholars.In order to explore the gap between humans and language models in informal reasoning ability, this paper constructs a Detective Reasoning Benchmark, which is an assembly of 1,200 questions gathered from accessible online resources, aims at evaluating the model's informal reasoning ability in real-life context.Considering the improvement of the model's informal reasoning ability restricted by the lack of benchmark, we further propose a Self-Question Prompt Framework that mimics human thinking to enhance the model's informal reasoning ability.The goals of self-question are to find key elements, deeply investigate the connections between these elements, encourage the relationship between each element and the problem, and finally, require the model to reasonably answer the problem.The experimental results show that human performance greatly outperforms the SoTA Language Models in Detective Reasoning Benchmark.Besides, Self-Question is proven to be the most effective prompt engineering in improving GPT-4's informal reasoning ability, but it still does not even surpass the lowest score made by human participants.Upon acceptance of the paper, the source code for the benchmark will be made publicly accessible.

Large pre-trained transformers are show-stealer in modern-day deep learning, and it becomes crucial to comprehend the parsimonious patterns that exist within them as they grow in scale. With exploding parameter counts, Lottery Ticket Hypothesis (LTH) and its variants, have lost their pragmatism in sparsifying them due to high computation and memory bottleneck of repetitive train-prune-retrain routine of iterative magnitude pruning (IMP) which worsens with increasing model size. This paper comprehensively studies induced sparse patterns across multiple large pre-trained vision and language transformers. We propose the existence of -- essential sparsity defined with a sharp dropping point beyond which the performance declines much faster w.r.t the rise of sparsity level, when we directly remove weights with the smallest magnitudes in one-shot without re-training. We also find essential sparsity to hold valid for N:M sparsity patterns as well as on modern-scale large language models (Vicuna-7B). We also present an intriguing emerging phenomenon of abrupt sparsification during the pre-training of BERT, i.e., BERT suddenly becomes heavily sparse in pre-training after certain iterations. Moreover, our observations also indicate a counter-intuitive finding that BERT trained with a larger amount of pre-training data tends to have a better ability to condense knowledge in comparatively relatively fewer parameters. Lastly, we investigate the effect of the pre-training loss on essential sparsity and discover that self-supervised learning (SSL) objectives trigger stronger emergent sparsification properties than supervised learning (SL). Our codes are available at \url{https://github.com/VITA-Group/essential_sparsity}.

Zero-shot sketch-based image retrieval (ZS-SBIR) is challenging due to the cross-domain nature of sketches and photos, as well as the semantic gap between seen and unseen image distributions. Previous methods fine-tune pre-trained models with various side information and learning strategies to learn a compact feature space that is shared between the sketch and photo domains and bridges seen and unseen classes. However, these efforts are inadequate in adapting domains and transferring knowledge from seen to unseen classes. In this paper, we present an effective ``Adapt and Align'' approach to address the key challenges. Specifically, we insert simple and lightweight domain adapters to learn new abstract concepts of the sketch domain and improve cross-domain representation capabilities. Inspired by recent advances in image-text foundation models (e.g., CLIP) on zero-shot scenarios, we explicitly align the learned image embedding with a more semantic text embedding to achieve the desired knowledge transfer from seen to unseen classes. Extensive experiments on three benchmark datasets and two popular backbones demonstrate the superiority of our method in terms of retrieval accuracy and flexibility.