1
AQuA：让量化研究 Agent 持续进化，也让回测结果经得起检验
自主研究系统的上限，不只取决于模型有多聪明，也取决于系统能否分辨什么是新证据，什么只是一次偶然的高分。
2
Towards an LLM-based method for quantifying the sexual content in song lyrics
用大模型量化雷鬼歌曲歌词的性内容，方法可复现，为音乐文本研究提供新工具
arXiv:2608.08885v1 Announce Type: cross Abstract: Reggaeton is one of the most widely consumed music genres in the world, and its lyrics are commonly …
3
Merit or networks? What decides where research is published
科学出版到底凭实力还是靠人脉？这篇新研究用方法突破传统难题，揭示背后的机制。
arXiv:2606.03763v1 Announce Type: cross Abstract: Does scientific publishing reward the quality of ideas or the advantage of connections? The question…
4
Quantifying the Impact of Translation Errors on Multilingual LLM Evaluation
量化翻译误差对多语言大模型评估的偏倚，揭示评估中的关键陷阱并改进可靠性。
arXiv:2605.24904v1 Announce Type: new Abstract: Machine-translated benchmarks are widely used to assess the multilingual capabilities of large languag…
5
Evaluating Large Language Models in a Complex Hidden Role Game
用隐藏角色游戏量化大模型欺骗潜力，研究设计新颖，评估视角独特。
arXiv:2605.22826v1 Announce Type: cross Abstract: Quantifying the deceptive potential of Large Language Models (LLMs) is critical for AI safety, yet d…