# AstroPT 对星系的认知如何校准 LLM 可解释性研究

- 来源：HuggingFace Daily Papers（社区热门论文）
- 发布时间：2026-08-23 08:00
- AIHOT 分数：55
- AIHOT 链接：https://aihot.virxact.com/items/cmt97tm0s0co9roly1nirqd8h
- 原文链接：https://arxiv.org/abs/2608.22614

## AI 摘要

研究者用基于数百万星系图像训练的 Transformer 模型 AstroPT 作为校准测试平台，检验概念在训练中涌现的顺序。探针分析发现，星系属性按已知难度固定顺序涌现：直接编码于像素的量（如波段星等）早期即可解码且位于浅层，而基于多波段/光谱推断的量（如红移、恒星形成率）出现更晚且位于深层。

## 正文

Interpretability research increasingly asks when concepts emerge during training and whether linear probes recover real structure, but in language models these claims are hard to validate because language offers little ground-truth ordering of concepts or relationships among them. We propose the use of astronomical ground truth through AstroPT, a transformer trained on millions of galaxy images, as a calibration testbed. AstroPT is an LLM-like model trained within a domain where the difficulty ordering of concepts and the relations among them are known in advance. Probing frozen representations across checkpoints, layers, model sizes, and objective choices, we find that galaxy properties emerge in a fixed order that tracks their known difficulty---quantities written almost directly into the pixels (band magnitude) become decodable early in training and shallow in the network, while multiband/spectra based and inferred quantities (such as redshift and specific star formation rate) emerge later and deeper. This order is invariant to our tested training objectives, and scales in magnitude but not in sequence with capacity. Our linear probe directions further recover the known physical structure among galaxy properties. Our findings suggest that astronomy offers a controlled sandbox for calibrating mechanistic interpretability methods we otherwise apply to LLMs blind.