Deep learning (DL) systems are increasingly applied to safety-critical domains such as autonomous driving cars. It is of significant importance ensure the reliability and robustness DL systems. Existing testing methodologies always fail include rare inputs in dataset exhibit low neuron coverage. In this paper, we propose DLFuzz, frst differential fuzzing framework guide exposing incorrect behaviors. DLFuzz keeps minutely mutating input maximize coverage prediction difference between original...

There has been renewed interests in the exploration of natural products (NPs) for drug discovery, and continuous investigations therapeutic claims mechanisms traditional herbal medicines. In-silico methods have employed facilitating these studies. These studies optimization in-silico algorithms NP applications can be facilitated by quantitative activity species source data NPs. A number databases collectively provide structural other information ∼470 000 NPs, including qualitative many but...

Abstract Drugs produce their therapeutic effects by modulating specific targets, and there are 89 innovative targets of first-in-class drugs approved in 2004–17, each with information about drug clinical trial dated back to 1984. Analysis the timelines these may reveal trial-speed differentiating features for facilitating target assessment. Here we present a comprehensive analysis all following earlier studies prospective prediction success drugs. Our confirmed literature-reported common...

Code reuse improves software development efficiency, however, vulnerabilities can be introduced inadvertently. Many existing works compute the code similarity based on CFGs to determine whether a binary function contains known vulnerability. Unfortunately, their performance in cross-platform search is challenged.

String similarity join is an important operation in data integration and cleansing that finds similar string pairs from two collections of strings. More than ten algorithms have been proposed to address this problem the recent decades. However, existing not thoroughly compared under same experimental framework. For example, some are tested only on specific datasets. This makes it rather difficult for practitioners decide which should be used various scenarios. To problem, paper we provide a...

Wireless medical cyber-physical systems are widely adopted in the daily practices of medicine, where huge amounts data sampled by wireless devices and sensors, is passed to decision support (DSSs). Many text-based guidelines have been encoded for work-flow simulation DSS automate health care based on those collected data. But some complex life-critical diseases, it highly desirable automatically rigorously verify temporal properties data, which brings new challenges current simulation-based...

The beneficial effects of functionally useful plants (e.g. medicinal and food plants) arise from the multi-target activities multiple ingredients these plants. knowledge collective molecular facilitates mechanistic studies expanded applications. A number databases provide information about targets various ingredients. More comprehensive is needed for broader classes landscapes individual plant's targets, regulated biological pathways, processes diseases. We therefore developed a new...

Identifying potentially vulnerable locations in a code base is critical as pre-step for effective vulnerability assessment; i.e., it can greatly help security experts put their time and effort to where needed most. Metric-based pattern-based methods have been presented identifying code. The former relies on machine learning cannot work well due the severe imbalance between non-vulnerable or lack of features characterize vulnerabilities. latter needs prior knowledge known vulnerabilities only...

Security of smart contracts has attracted increasing attention in recent years. Many researchers have devoted themselves to devising testing tools for vulnerability detection. Each published tool demonstrated its effectiveness through a series evaluations on their own experimental scenarios. However, the inconsistency evaluation settings such as different data sets or performance metrics, may result biased conclusion.

Taint analysis assists fuzzers in solving complex fuzzing constraints by inferring the influencing input bytes. Execution paths real-world programs often reach loops, where these loops can be visited and recorded multiple times. Conventional taint techniques experience difficulties when distinguishing between occurrences of same constraint. In this paper, we propose PATA, a fuzzer that implements path-aware analysis, i.e. one distinguishes variable based on execution path information. PATA...

S v -ZIS/N–TiO 2 nanoarrays produced a photocurrent density of 4.9 mA cm −2 at an external potential 1.23 V vs. RHE, and achieved PEC H evolution rate 49.59 mmol g −1 h . The IPCE the photoanode device is ∼57.9% 350 nm ∼7.3% 400 nm.

This paper proposes LATTE, the first static binary taint analysis that is powered by a large language model (LLM). LATTE superior to state of art (e.g., Emtaint, Arbiter, Karonte) in three aspects. First, fully automated while prior analyzers need rely on human expertise manually customize propagation rules and vulnerability inspection rules. Second, significantly effective detection, demonstrated our comprehensive evaluations. For example, has found 37 new bugs real-world firmware, which...

Smart contracts, as a promising and powerful application on the Ethereum blockchain, have been growing rapidly in past few years. Since they are highly vulnerable to different forms of attacks, their security becomes top priority. However, existing auditing techniques either limited fnding vulnerabilities (rely pre-defned bug paterns) or very expensive program analysis), thus insufcient for Ethereum.

Mutation-based fuzzing is a widely used software testing technique for bug and vulnerability detection, the performance greatly affected by quality of initial seeds effectiveness mutation strategy. In this paper, we present SAFL1, an efficient tool augmented with qualified seed generation coverage-directed mutation. First, symbolic execution in lightweight approach to generate seeds. Valuable explore directions are learned from seeds, thus later process can reach deep paths program state...

Researchers have proposed many optimizations to improve the efficiency of fuzzing, and most optimized strategies work very well on their targets when running in single mode with instantiating one fuzzer instance. However, real industrial practice, fuzzers run parallel multiple instances, those unfortunately fail maintain improvements.

Ethereum Virtual Machine (EVM) is the run-time environment for smart contracts and its vulnerabilities may lead to serious problems ecology. With lots of techniques being continuously developed validation contracts, testing EVM remains challenging because special test input format absence oracles. In this paper, we propose EVMFuzzer, first tool that uses differential fuzzing technique detect EVM. The core idea generate seed feed them target benchmark EVMs, so as find many inconsistencies...

Deep Learning (DL) system has been widely used in many critical applications, such as autonomous vehicles and unmanned aerial vehicles. However, their security is threatened by backdoor attack, which achieved adding artificial patterns on specific training data. Existing attack methods normally poison the data using a patch, they can be easily detected existing detection methods. In this work, we propose Adversarial Backdoor, utilizes Targeted Universal Perturbation (TUAP) to hide anomalies...

Due to extensive bioprospecting efforts of the past and technology factors, there have been questions about drug discovery prospect from untapped species. We analyzed recent trends approved drugs derived previously species, which show no sign drug-productive species being near extinction suggest high probability deriving new in existing families clusters. Case histories recently reveal useful strategies for scaffolds pharmacophores natural product leads these New technologies such as cryptic...

Fuzzing is widely used for software vulnerability detection. There are various kinds of fuzzers with different fuzzing strategies, and most them perform well on their targets. However, in industry practice empirical study, the performance generalization ability those well-designed strategies challenged by complexity diversity real-world applications. In this paper, inspired idea ensemble learning, we first propose an approach EnFuzz, that integrates multiple to obtain better than any...

Today's system-on-chip and distributed systems are commonly equipped with multiple clocks. The key challenge in designing such is that two situations have to be captured evaluated a single framework. first the heterogeneous control-oriented data-oriented behaviors within one clock domain, second asynchronous communications between domains. In this paper, we propose use timed automata synchronous dataflow model dynamic of multiclock train-control system, multiprocessor architecture for...

Fuzz testing has helped security researchers and organizations discover a large number of vulnerabilities. Although it is efficient widely used in industry, hardly any empirical studies experience exist on the customization fuzzers to real industrial projects. In this paper, collaborating with engineers from Huawei, we present practice adapting fuzz proprietary message middleware named libmsg, which responsible for transfer entire distributed system department. We main obstacles coming...

As a problem-solving method, neural networks have shown broad success for medical applications, speech recognition, and natural language processing. Current hardware implementations of exhibit high energy consumption due to the intensive computing workloads. This paper proposes methodology design an energy-efficient network that effectively exploits computation reuse opportunities. To do so, we use Bloom filters (BFs) by tightly integrating them with units. BFs store recall frequently...

Industrial Control System (ICS) protocols play an essential role in building communications among system components. Recently, many severe vulnerabilities, such as Stuxnet and DragonFly, exposed ICS have affected a wide distribution of devices. Therefore, it is vital importance to ensure their correctness. However, the vulnerability detection efficiency traditional techniques fuzzing challenged by complexity diversity protocols.In this paper, we propose equip protocol with coverage-guided...