Open-Ended Generation Steering

Task of steering LLM free-text responses toward psychological constructs; the primary evaluation regime where injections outperform prompting

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

geometry-based steeringconcept0.756
Paradigm of finding the right geometry (manifold) for principled control.
direction-based steeringconcept0.751
Paradigm of finding the right direction in activation space (e.g., linear steering).
Endogenous Steering Resistanceconcept0.746
The central phenomenon introduced by this paper: inference-time recovery from irrelevant activation steering in LLMs
Activation Steeringmethod0.746
Causal intervention technique: edit NLA explanation, reconstruct via AR, use difference as steering vector to manipulate model behavior.
Model Steeringconcept0.743
Using interventions to guide model generation behavior, e.g., adding sentiment vectors at inference time
Interpretability-driven steeringconcept0.731
General approach of using interpretability feedback to steer model generation.
Representation Steeringconcept0.729
Parent concept; the practice of controlling neural network outputs by manipulating internal representations.
steering (intervention on internals)concept0.723
General technique of modifying activations to control model behavior.