Tech Meridian ← LIVE FEED
RU

RESEARCH · RESEARCH · #629

Dynamically Scaled Activation Steering (DSAS) adaptively modulates steering in generative models

DSAS is a method-agnostic framework that decouples when to steer from how to steer by computing context-dependent scaling factors to modulate the strength of existing activation-steering transformations across layers and inputs. The authors report that DSAS improves the trade-off between toxicity mitigation and utility preservation, adds minimal compute overhead, improves interpretability by highlighting tokens that require steering, and can be jointly optimized with steering functions; the paper was accepted to the UniReps workshop at NeurIPS 2025 and the authors say code will be made available on GitHub.

KEY POINTS

  1. DSAS is a method-agnostic framework that decouples when to steer from how to steer by computing context-dependent scaling factors to modulate the strength of existing activation-steering transformations across layers and inputs.
  2. The authors report that DSAS improves the trade-off between toxicity mitigation and utility preservation, adds minimal compute overhead, improves interpretability by highlighting tokens that require steering, and can be jointly optimized with steering functions; the paper was accepted to the UniReps workshop at NeurIPS 2025 and the authors say code will be made available on GitHub.
  3. By applying steering only when and where needed, DSAS can improve safety-utility trade-offs and make steering behavior more interpretable across model types.

WHY IT MATTERS

By applying steering only when and where needed, DSAS can improve safety-utility trade-offs and make steering behavior more interpretable across model types.

SOURCES & TIMELINE

1