ScholarPulse 日报 2026-08-17

2026-08-17 学术简报：2 篇。LLM-modernized科学软件内核在故障响应测试中与原始代码完全一致，验证了现代化的可靠性。

今日速览

序号	标题	来源	日期	主题	推荐等级
1	Validating LLM-Modernized Scientific Software Through Differential Fault Injection	arXiv	2026-08-14	AI-Agent	高
2	Participatory Moral AI Is Not Neutral: The Invisible Hand of Developers	arXiv	2026-08-14	AI-Agent	高

重点论文与技术动态

1. Validating LLM-Modernized Scientific Software Through Differential Fault Injection

来源：arXiv
日期：2026-08-14
作者/机构：Evan Coleman, Yuzhong Shen, Masha Sosonkina, Peng Xu
主题标签：AI-Agent, arXiv
推荐等级：高
分类：cs.DC, cs.SE

一句话结论

LLM-modernized科学软件内核在故障响应测试中与原始代码完全一致，验证了现代化的可靠性。

核心内容

LLM现代化的科学软件验证常侧重名义执行，忽略故障、扰动和精度降低下的行为测试。
2200多次运行中，瞬态故障吸收成本与收缩模型匹配（预测斜率0.74和1.49迭代/比特，实测0.82和1.50），持久扰动使最终能量误差每增加一位减半。
原始与现代内核在200次配对注入中完全一致，精度降低下暴露相位依赖并行死锁和虚假收敛。

方法与数据

方法：在GAMESS的12个站点应用相同确定性故障，隔离转换的积分内核。
摘要未明确：具体故障类型或更多实验参数细节。

价值判断

值得关注：LLM现代化的验证需包含故障响应测试，避免名义执行的局限，确保实际场景可靠性。
可复用点：差分故障注入方法可直接推广至其他LLM-modernized科学软件的验证流程。
局限/待核查：精度降低下暴露的死锁和虚假收敛现象需进一步分析验证。

摘要

Large language model (LLM) agents are increasingly used to modernize the legacy Fortran underlying production scientific software, but validation of these transformations emphasizes nominal executions and may not test whether a modernization preserves the original code's response to faults, perturbations, and reduced precision. We present a differential fault-injection validation method: a harness instruments the shared self-consistent-field driver of GAMESS at twelve sites and applies identical, deterministic faults to the original and LLM-modernized implementations, isolating the converted integral kernels. Across more than 2,200 runs, transient-fault absorption costs match a contraction-based model (predicted slopes 0.74 and 1.49 iterations per bit; measured 0.82 and 1.50), persistent perturbations halve final-energy error per additional bit, and the campaigns expose phase-dependent parallel deadlocks and false convergence under reduced precision. The original and modernized kernels agree in all 200 paired injections, and a measurement-guided synchronization change composes with the modernization, matching in all 40 pairs.

2. Participatory Moral AI Is Not Neutral: The Invisible Hand of Developers

来源：arXiv
日期：2026-08-14
作者/机构：Taenyun Kim, Edyta Bogucka, Daniele Quercia
主题标签：AI-Agent, arXiv
推荐等级：高
分类：cs.AI

一句话结论

投票式道德AI对齐无法仅通过聚合实现公平透明，必须审计披露管道各阶段。

一段话。
开发者在道德AI获取管道中对特征范围、投票者抽样和问题框架的决策常被忽视为技术细节，实际影响偏好生成，导致道德AI非中立。

核心内容

特征范围界定在AI肾脏分配、AI代理模拟缺席员工等上下文中不通用，特征模式不可跨领域迁移。
政治意识形态影响约三分之一特征的偏好，差异方向可反转，影响聚合偏好结果。
问题措辞能缩小或扩大意识形态差距达一个量级，并改变道德基础关联方式。

方法与数据

两阶段研究，样本量809人，覆盖AI肾脏分配、AI代理模拟缺席员工及生成式AI描绘逝者三个场景。
摘要未明确实验设计细节。

价值判断

值得关注：开发者隐性决策导致道德AI非中立，需透明化处理以避免偏见。
可复用点：研究框架可推广至其他AI道德决策系统评估。
局限/待核查：摘要未明确。

摘要

As AI systems make more morally loaded decisions across society, one response has been moral preference elicitation. In this approach, researchers poll participants on hypothetical dilemmas and use the aggregated votes to train a policy that an AI model then applies at scale. Before any vote is cast, developers make three key choices in the moral AI elicitation pipeline: feature scoping, voter sampling, and question framing. In other words, they decide which features go to a vote, which voters to include, and how to present the question. These choices are often opaque, undocumented, and treated as technical details rather than normative ones. We examine each of these choices within a common empirical study and show that each can shape the preferences produced by moral AI elicitation. Across two phases (N = 809) in three deployment contexts (i.e., AI kidney allocation, AI agents simulating absent workers, and generative AI depictions of the deceased), we examine the three main stages of the moral AI elicitation pipeline. First, morally relevant features shift across contexts. This suggests that feature schemas should not be assumed to transfer across deployment domains. Second, preferences differ by political ideology for roughly one-third of features, with some differences reversing direction. The ideological composition of the voter pool can therefore affect the resulting aggregated preference profile. Third, the wording of the elicitation question can narrow or widen ideological gaps by up to a full scale point. The framing conditions also change how moral foundations are associated with participants' judgments. Taken together, these findings suggest that voting-based alignment cannot deliver fair or transparent AI by aggregation alone; at minimum, each stage of the moral AI elicitation pipeline should be audited and disclosed.