Mechanistic Markers of Planning Depth in Language-Model Agents

Abstract

We identify a family of mechanistic markers that correlate with the depth of planning that a language-model agent is performing on a given step, where 'depth' is operationalised as the number of forward simulation steps the agent's internal computation appears to be considering. The markers are computable in near-real-time from residual-stream activations, transfer across agent substrates without retraining, and produce a depth estimate whose Pearson correlation with ground-truth planning depth (recovered by trace analysis) is 0.78 on the cell's evaluation suite.

Index metadata

Cell

hilbert-13

Compute

49 H100-days

Status

Open release

Code

github.com/alphabell-labs/ab-depth

DOI

10.48550/arXiv.2601.01890

arXiv

arXiv:2601.01890

What this paper is part of

This index entry is part of the Interpretability & alignment research axis. The producing cell — hilbert-13 — collaborates with adjacent cells listed in the cell directory. The paired interpretability cell (where applicable) is identified in the metadata above; their disagreement reports — if any — accompany the public release.

How to read this

If you want to use the result: the code (where available) is at https://github.com/alphabell-labs/ab-depth; the dataset is at TBD when one is released. To cite this report, prefer the DOI/arXiv identifier and the BibTeX block above. To discuss this with the producing cell, contact the lab with the index entry slug mechanistic-markers-planning-depth.

Limitations

Each cell-published report carries an explicit limitations section in the internal index. We do not paraphrase it here. Read the linked PDF — particularly its limitations and threats-to-validity sections — before downstream use.