Measuring Concept Content in Text from LLM Activations: ESG Evidence from Concept Vectors and Linear Probes

Luc Hazenoot, Zhaochun Ren and Amirhossein Zohrehvand

Papers from arXiv.org

Abstract: Recent work shows that Large Language Models (LLMs) encode more information in their activations than they show in their responses. We ask whether the information contained in activations could be used as a source of measurement for the presence of concepts. We compare two algorithms for extraction of concepts, i.e., Recursive Feature Machine (RFM) algorithm and linear probing against an embedding baseline, surface baselines, and the model's own response using a human-annotated Environmental, Social and Governance (ESG) dataset of financial texts. The best linear probe ranges from 0.6 to 3.3 percentage points below fine-tuned domain classifiers. The probes also outscore the model's own response in all comparisons. Furthermore, the probes outperform the RFM concept vectors in all accuracy comparisons. In a synthetic mixture test, their mean-pooled score tracks the number of concept sentences in a paragraph better than the concept-vector score. Put together, a linear probe on a frozen LLM is a cost-effective and generalizable alternative to fine-tuning a domain classifier.

Date: 2026-08, Revised 2026-10
New Economics Papers: this item is included in nep-big and nep-cmp
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://arxiv.org/pdf/2608.07208 Latest version (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:arx:papers:2608.07208

Access Statistics for this paper

More papers in Papers from arXiv.org
Bibliographic data for series maintained by arXiv administrators ().