# Can Test-Time Memory Evolution Outperform Test-Time Training for Scientific Discovery?

**Published:** October 26, 2026
**Author:** Yepeng Liu, Shengmai Chen, Xuandong Zhao, Chengzhi Liu, Dawn Song, Xin Eric Wang, Yuheng Bu
**Source:** arXiv — cs.AI
**Language:** en-US
**HTML:** https://aiqazaqstan.com/en/articles/can-test-time-memory-evolution-outperform-test-time-training-for-scienti

arXiv:2610.32046v1 Announce Type: new Abstract: Recent \emph{test-time training} (TTT) frameworks have enabled open-weight large language models (LLMs) to outperform inference-only frameworks powered by frontier models on scientific discovery tasks. However, TTT introduces…

> From the original source: **arXiv — cs.AI**
> [Read full article at arXiv — cs.AI](https://arxiv.org/abs/2610.32046)

**Tags:** #arXiv

---

*AI Qazaqstan — non-commercial informational portal about AI in Kazakhstan. Machine translations and summaries are AI-generated.*