Conceptors for Semantic Steering

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Triantafyllopoulos, Ilias, Cho, Young-Min, Tao, Ren, Miao, Miranda Muqing, Rai, Sunny, Ungar, Lyle, Guntuku, Sharath Chandra, Ryant, Neville, Sedoc, João
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917464356421632
author Triantafyllopoulos, Ilias
Cho, Young-Min
Tao, Ren
Miao, Miranda Muqing
Rai, Sunny
Ungar, Lyle
Guntuku, Sharath Chandra
Ryant, Neville
Sedoc, João
author_facet Triantafyllopoulos, Ilias
Cho, Young-Min
Tao, Ren
Miao, Miranda Muqing
Rai, Sunny
Ungar, Lyle
Guntuku, Sharath Chandra
Ryant, Neville
Sedoc, João
contents Activation-based steering provides control of LLM behavior at inference time, but the dominant paradigm reduces each concept to a single direction whose geometry is left largely unexamined. Rather than selecting a single steering direction, we use conceptors: soft projection matrices estimated from activations pooled across both poles of a bipolar concept, which preserve the concept's full multidimensional subspace. A geometric analysis shows the bipolar subspace strictly subsumes the single-vector baseline. We further show that the conceptor quota provides a parameter-free layer-selection diagnostic, predicting concept separability with Pearson correlations up to r=0.96 across three instruction-tuned models and three semantic dimensions. Beyond selection, conceptors admit a closed-form Boolean algebra (AND, OR, NOT): we evaluate conceptor compositionality on thematically related sub-concepts. Across a systematic five-axis design-space evaluation, conceptors match or outperform additive baselines at layers where concept subspaces are multi-dimensional while producing substantially fewer degenerate outputs. Conceptor steering is a geometrically principled, compositional, and practically safer alternative to single-direction steering from a limited number of contrastive pairs.
format Preprint
id arxiv_https___arxiv_org_abs_2605_04980
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Conceptors for Semantic Steering
Triantafyllopoulos, Ilias
Cho, Young-Min
Tao, Ren
Miao, Miranda Muqing
Rai, Sunny
Ungar, Lyle
Guntuku, Sharath Chandra
Ryant, Neville
Sedoc, João
Machine Learning
Computation and Language
Activation-based steering provides control of LLM behavior at inference time, but the dominant paradigm reduces each concept to a single direction whose geometry is left largely unexamined. Rather than selecting a single steering direction, we use conceptors: soft projection matrices estimated from activations pooled across both poles of a bipolar concept, which preserve the concept's full multidimensional subspace. A geometric analysis shows the bipolar subspace strictly subsumes the single-vector baseline. We further show that the conceptor quota provides a parameter-free layer-selection diagnostic, predicting concept separability with Pearson correlations up to r=0.96 across three instruction-tuned models and three semantic dimensions. Beyond selection, conceptors admit a closed-form Boolean algebra (AND, OR, NOT): we evaluate conceptor compositionality on thematically related sub-concepts. Across a systematic five-axis design-space evaluation, conceptors match or outperform additive baselines at layers where concept subspaces are multi-dimensional while producing substantially fewer degenerate outputs. Conceptor steering is a geometrically principled, compositional, and practically safer alternative to single-direction steering from a limited number of contrastive pairs.
title Conceptors for Semantic Steering
topic Machine Learning
Computation and Language
url https://arxiv.org/abs/2605.04980