arXiv subjects

Sun Jun

Publications and source records attributed to Sun Jun.

4 recordsLinked to original sources

arXiv2026

The 2026 Singapore Consensus on Global AI Safety Research Priorities

Frontier AI capabilities and autonomy are advancing rapidly. A growing number of real-world incidents make a trusted AI ecosystem essential to embracing AI with confidence. The 2026 Singapore Consensus is an outcome of the second International Scientific Exchange on AI Safety, bringing together over 100 contributors spanning 13 countries from frontier developers, government safety institutes, academia, and civil society. Building on the 2025 report, it presents a global understanding of technical AI safety research problems of top priority, now with a dedicated focus on societal resilience and on managing the risks of increasingly autonomous AI agents.

cs.CY↗

arXiv2021

Whether the Health Care Practices For the Patients With Comorbidities Have Changed After the Outbreak of COVID-19; Big Data Public Sentiment Analysis

After the pandemic of SARS-CoV-2, it has influenced the health care practices around the world. Initial investigations indicate that patients with comorbidities are more fragile to this SARS-CoV-2 infection. They suggested postponing the routine treatment of cancer patients. However, few meta-analyses suggested evidences are not sufficient to hold the claim of the frailty of cancer patients to COVID-19, and they are not in favour of shelving the scheduled procedures. There are recent studies in which medical professionals, according to their competence, are referring to change the routine practices on how to manage the applicable therapeutic resources judiciously to combat this vital infection. This is a different study that reveals the cancer patients' viewpoint about how health care practices have been changed in their opinion during this pandemic year? Are they satisfied with their treatment or not? To serve the purpose, we gathered more than 60000 relevant tweets from Twitter to analyse the sentiment of cancer patients around the world. Our findings demonstrate that there is a surge in argument about cancer and its treatment after the outbreak of COVID-19. Most of the tweets are reasonable (52.6%) compared to the negative ones (24.3). We developed polarity and subjectivity distribution to better recognise the positivity/negativity in the sentiment. Results reveal that the polarity range of positive tweets is within the range of 0 to 0.5. Which means the tendency in the tweets is not so much positive but surely not negative. It is a piece of modest statistical evidence in support of how natural language processing (NLP) can be accepted to better understand the patient's behaviour in real-time, and it may facilitate the medical professional to make better decision to organise the routine management of cancer patients.

cs.CY↗

arXiv2019

Systematic Classification of Attackers via Bounded Model Checking

In this work, we study the problem of verification of systems in the presence of attackers using bounded model checking. Given a system and a set of security requirements, we present a methodology to generate and classify attackers, mapping them to the set of requirements that they can break. A naive approach suffers from the same shortcomings of any large model checking problem, i.e., memory shortage and exponential time. To cope with these shortcomings, we describe two sound heuristics based on cone-of-influence reduction and on learning, which we demonstrate empirically by applying our methodology to a set of hardware benchmark systems.

cs.CR↗

arXiv2018

Quantifying Attacker Capability Via Model Checking Multiple Properties (Extended Version)

This work aims to solve a practical problem, i.e., how to quantify the risk brought upon a system by different attackers. The answer is useful for optimising resource allocation for system defence. Given a set of safety requirements, we quantify the attacker capability in terms of the set of safety requirements an attacker can compromise. Given a system (in the presence of an attacker), model checking it against each safety requirement one by one is expensive and wasteful since the same state space is explored many times. We thus propose model checking multiple properties efficiently by means of coalgebraic model checking using enhanced coinduction techniques. We apply the proposed technique to a real-world water treatment system and the results show that our approach can effectively reduce the effort required for model checking.

cs.LO↗