As the least researched component of holistic Life Cycle Sustainability Assessment (LCSA), Social Life Cycle Assessment (S-LCA) continues to face significant methodological fragmentation—hindering the integration of social dimensions in sustainability research. Despite the methodological consolidation efforts provided by the landmark 2020 United Nations Environment Programme (UNEP) Guidelines, several challenges remain. To lay a foundation for greater coherence in S-LCA, this study conducts a critical review of 30 case studies published since the 2020 UNEP Guidelines to evaluate both emerging practices and enduring limitations. From this, four key methodological gaps are identified: (1) lack of transparency for rigor and replicability, (2) macro-scale approaches with generic data inhibiting contextualized analysis, (3) data availability limitations, and (4) need for more effective application of qualitative methods. To address these challenges, the review offers future directions, including enhanced documentation protocols, adoption of micro-scale boundaries enriched with high-resolution data, and shifting application of databases. Additionally, its discussion highlights the importance of high-fidelity data and collaboration with corporations, alongside qualitative method applications adherent to social science principles. These insights inform a flexible framework that builds from existing standards and guidelines to support rigorous, context-specific S-LCA implementation across diverse sectors and micro-scale applications.
Large Language Models (LLMs) have emerged as a new information channel. Meanwhile, one critical but under-explored question is: Is it possible to bypass the safety alignment and inject harmful information into LLMs stealthily? In this paper, we propose to reformulate knowledge editing as a new type of safety threat for LLMs, namely Editing Attack, and conduct a systematic investigation with a newly constructed dataset EditAttack. Specifically, we focus on two typical safety risks of Editing Attack including Misinformation Injection and Bias Injection. For the first risk, we find that editing attacks can inject both commonsense and long-tail misinformation into LLMs, and the effectiveness for the former one is particularly high. For the second risk, we discover that not only can biased sentences be injected into LLMs with high effectiveness, but also one single biased sentence injection can degrade the overall fairness. Then, we further illustrate the high stealthiness of editing attacks. Our discoveries demonstrate the emerging misuse risks of knowledge editing techniques on compromising the safety alignment of LLMs and the feasibility of disseminating misinformation or bias with LLMs as new channels.
In my years as a corporate law academic, I've been surprised at the paucity of interactions between those who study corporate law and those who "do" it. Those who do corporate law include not only lawyers who practice corporate law and judges who make and interpret corporate law, but also bankers and business people who are integrally involved in corporate transactions. The Theory Informs Business Practice symposium aimed to increase interaction among all of these actors in the corporate law world.
Time series anomaly detection (TSAD) has been a long-standing pillar problem in Web-scale systems and online infrastructures, such as service reliability monitoring, system fault diagnosis, and performance optimization. Large language models (LLMs) have demonstrated unprecedented capabilities in time series analysis, the potential of multimodal LLMs (MLLMs), particularly vision-language models, in TSAD remains largely under-explored. One natural way for humans to detect time series anomalies is through visualization and textual description. It motivates our research question: Can multimodal LLMs perform time series anomaly detection? Existing studies often oversimplify the problem by treating point-wise anomalies as special cases of range-wise ones or by aggregating point anomalies to approximate range-wise scenarios. They limit our understanding for realistic scenarios such as multi-granular anomalies and irregular time series. To address the gap, we build a VisualTimeAnomaly benchmark to comprehensively investigate zero-shot capabilities of MLLMs for TSAD, progressively from point-, range-, to variate-wise anomalies, and extends to irregular sampling conditions. Our study reveals several key insights in multimodal MLLMs for TSAD. Built on these findings, we propose a MLLMs-based multi-agent framework TSAD-Agents to achieve automatic TSAD. Our framework comprises scanning, planning, detection, and checking agents that synergistically collaborate to reason, plan, and self-reflect to enable automatic TSAD. These agents adaptively invoke tools such as traditional methods and MLLMs and dynamically switch between text and image modalities to optimize detection performance.
Randomized quasi-Monte Carlo (RQMC) methods estimate the mean of a random variable by sampling an integrand at $n$ equidistributed points. For scrambled digital nets, the resulting variance is typically $\tilde O(n<^>{-\theta })$, where $\theta \in [1,3]$ depends on the smoothness of the integrand and $\tilde O$ neglects logarithmic factors. While RQMC can be far more accurate than plain Monte Carlo (MC), it remains difficult to get confidence intervals on RQMC estimates. We investigate some empirical Bernstein confidence intervals (EBCIs) and hedged betting confidence intervals (HBCIs), both from Waudby-Smith and Ramdas (2024, J. Roy. Statist. Soc. B, 86, 1-27), when the random variable of interest is subject to known bounds. When there are $N$ integrand evaluations partitioned into $R$ independent replicates of $n=N/R$ RQMC points, and the RQMC variance is $\varTheta (n<^>{-\theta })$, then an oracle minimizing the width of a Bennett confidence interval would choose $n =\varTheta (N<^>{1/(\theta +1)})$. The resulting intervals have a width $\varTheta (N<^>{-\theta /(\theta +1)})$. Our empirical investigations had optimal values of $n$ grow slowly with $N$, HBCI intervals that were usually narrower than the EBCI ones and optimal values of $n$ for HBCI that were equal to or smaller than the ones for the oracle.
