1
Stress Testing Unlearning Algorithms
系统检验机器遗忘算法在极端压力下的鲁棒性，为隐私保护研究划出新基准。
arXiv:2608.22527v1 Announce Type: new Abstract: Recently, machine unlearning, the removal of specific training data influence from a model, has gained…
2
Decoding-Level Taboo: A Diagnostic Stress Test for LLM Robustness
用解码级禁忌测试，揭示大模型在边界情境下的真实鲁棒性，为AI安全诊断提供新思路。
arXiv:2608.09900v1 Announce Type: new Abstract: Large language model evaluations typically focus on performance under nominal conditions, creating an …
3
Show HN: Grinta –local-first AI coding agent that passed 106-minute stress test
Grinta本地优先AI编程代理，自主运行4.5小时，完成373次工具调用与39项测试，无需人工介入。
Article URL: https://github.com/josephsenior/Grinta-Coding-Agent Comments URL: https://news.ycombinator.com/item?id=48808609 Points: 1 # Comments: 0
4
Patronus AI lands $50M to build ‘digital worlds’ that stress-test AI agents
5000万美元融资，用“数字世界”模拟超长运转压力，专测AI代理可靠性。
Agent testing startup Patronus AI, founded by former Meta AI researchers, is experienced nearly insatiable demand, its investor says.
5
TamperBench: Systematically Stress-Testing LLM Safety Under Fine-Tuning and Tampering
系统性压力测试大模型在微调与篡改下的安全性，为LLM红队评估提供新基准。
arXiv:2602.06911v2 Announce Type: replace-cross Abstract: As increasingly capable open-weight large language models (LLMs) are deployed, improving the…
6
How to Stress-Test LLM Judges Fairly
提出固定预算与集群感知标准，公平测试LLM作为裁判的可靠性，多跳RAG压力测试新方法。
Article URL: https://www.alphaxiv.org/abs/2605.27789 Comments URL: https://news.ycombinator.com/item?id=48322235 Points: 2 # Comments: 0
7
MemFail: Stress-Testing Failure Modes of LLM Memory Systems
用压力测试方法系统揭示LLM记忆系统的关键失效模式，为改进大模型记忆能力提供实证依据。
arXiv:2605.26667v1 Announce Type: new Abstract: Large language model (LLM) agents increasingly rely on external memory systems to remain consistent ac…
8
Wake-Up Call: Why AI Safety Guardrails Break Under Pressure
六大模型压力测试揭示AI安全护栏在持续追问下会失效，值得关注
This is a submission for the Google I/O Writing Challenge This is a submission for the Google I/O Writing Challenge We treat AI safety as a static sta…
9
AI Edits a Classical Chinese Paper: Multi-Model Stress Test
多模型协同编辑古典中文论文，一场探索AI处理复杂文本能力的压力测试。
Article URL: https://zenodo.org/records/20343571 Comments URL: https://news.ycombinator.com/item?id=48235251 Points: 1 # Comments: 0
10
MANTA: Multi-turn Assessment for Nonhuman Thinking & Alignment
多轮对话评估揭示AI在动物福利对齐上的隐蔽失败，压力下模型会背离初始立场。
arXiv:2605.16301v1 Announce Type: cross Abstract: Single-turn benchmarks such as AnimalHarmBench (AHB) have established important baselines for measur…