DimSteer: Steering LLM Authoring with Automatically Discovered Stylistic Controls

Ajit Mallavarapu; Ziwei Gu

cs.HCOct 3, 2026

DimSteer: Steering LLM Authoring with Automatically Discovered Stylistic Controls

Authors: Ajit Mallavarapu, Ziwei Gu

Organizations: Harvard University, USA

Abstract

Large language model writing interfaces often make users steer outputs by repeatedly articulating desired changes in natural language. Yet writers may recognize useful stylistic directions only after seeing alternatives, making revision recall-heavy. We present DimSteer, an authoring interface that samples prompt-local completions, discovers high-variance activation-space axes of variation, labels them, and exposes them as sliders with pole previews, diff comparison, and reset controls. Users can manipulate discovered dimensions, reducing the need to reformulate prompts for each stylistic adjustment. In a within-subjects study with 16 participants against a matched prompt-only baseline, DimSteer reduced mental demand, effort, and frustration while preserving comparable perceived success. Participants valued the surfaced dimensions, yet 15 of 16 disagreed that they would have thought to request the same changes in a prompt. Results suggest prompt-local controls can shift LLM authoring from recall-based prompting toward recognition-based exploration and direct manipulation, while preserving prompting for open-ended edits.

Figures & tables

Figure 1. DimSteer shifts LLM interaction from recall-based prompting to direct manipulation. (1) Standard chat interfaces require users to articulate the exact changes needed to steer text. (2) DimSteer instead surfaces automatically discovered dimensions of generation variation as continuous controls, allowing users to explore variation through recognition and direct manipulation. (3) When discovered controls are insufficient, users can define custom dimensions to extend the machine-proposed control space. A three-stage overview of DimSteer showing a transition from recall-based prompting, to automatically discovered semantic sliders, to user-defined repair and extension of the control space.

Figure 2. Interface A: Prompt-Only Baseline. Participants revised drafts using standard natural-language instructions.

Figure 3. The DimSteer Interface. (A) Each discovered slider card includes (A1) current steering coefficient, (A2) PCA variance tag, and (A3) directional pole text previews. (B) Reset returns steering to the baseline. (C) Custom-dimension field lets users add missing controls. (D) Diff toggle reveals changes from the baseline draft.

Panel of Figure 4 — Figure 4. DimSteer implementation pipeline. The system samples prompt-local completions, extracts layer-21 activations, identifies PCA axes, labels and orients them with an LLM judge, and applies the resulting vectors through slider-based activation steering.

Prompt	Task prompt	Surfaced dimensions
P1	Describe the chaotic energy of a packed night market in a bustling city.	Vivid, Sensational, Atmosphere
P2	Describe a sudden thunderstorm hitting a quiet, open field during a hot afternoon.	Vividness, Sensory, Dissonance
P3	Write a brief, encouraging announcement message for a team channel celebrating a successful project launch.	Formality, Enthusiasm, Tone
P4	Write a brief, urgent Slack message asking teammates to hop onto a bridge call to fix a sudden production crash.	Urgency, Formality, Directness

Table 1. Shared study prompts and surfaced dimensions.

Measure	$C_{PROMPT}$	$C_{STEER}$
Mean task duration	5.72 min	3.22 min
Median task duration	5.70 min	3.20 min
Task duration range	5.3–6.2 min	3.0–3.5 min
Median revision actions	3.5 reprompts	4.5 slider changes
Revision action range	3–4	2–7
Diff view used	11/16	13/16 participants

Table 2. Descriptive interaction-log summaries across participants and tasks.

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Prompt	PCA-discovered controls	GPT5.6-luna-generated controls
P1	Vivid; Sensational; Atmosphere	Immersiveness; Sensory Intensity; Pacing
P2	Vividness; Sensory; Dissonance	Dramatic Contrast; Sensory Detail; Atmospheric Tension
P3	Formality; Enthusiasm; Tone	Warmth; Enthusiasm; Formality
P4	Urgency; Formality; Directness	Directness; Reassurance; Urgency

Appendix

Table 3. Control sets shown in the supplementary control-source evaluation.

Language Model Steering

Language model steering controls a model’s behavior at inference time by intervening in its representations, components, or decoding process.

See all 279 papers

Steer Like the LLM: Activation Steering that Mimics Prompting
May 5, 2026Geert Heyman, Frederik VandeputteLanguage Model Steering LLM Prompting
Nokia Bell Labs, Belgium.
From Words to Widgets for Controllable LLM Generation
Apr 13, 2026Chao Zhang, Yiren Liu, Lunyiu Nie +3LLM Prompting Controllable Text Generation
Cornell University Ithaca, NY, USA · University of Illinois Urbana-Champaign Champaign, IL, USA · The University of Texas at Austin Austin, TX, USA +1
Can Activation Steering Capture Multidimensional Authorship Style?
Sep 4, 2026Hieu Tran, Calvin Bao, Marine CarpuatLanguage Model Steering Activation Steering
University of Maryland, College Park

May 5, 2026J/K move · Enter open · S save

Steer Like the LLM: Activation Steering that Mimics Prompting

Geert Heyman, Frederik Vandeputte

Nokia Bell Labs, Belgium.

Language Model Steering LLM Prompting Causal Interventions in Language Models Activation Steering

Open paper AskPDF arXiv

Large language models can be steered at inference time through prompting or activation interventions, but activation steering methods often underperform compared to prompt-based approaches. We propose a framework that formulates prompt steering as a form of activation steering and investigates whether distilling successful prompt steering behavior into simpler, interpretable models can close this gap. Our analysis reveals that popular activation steering methods are not faithful to the mechanics of prompt steering, which applies strong interventions on some tokens while barely affecting others. Based on these insights, we introduce Prompt Steering Replacement (PSR) models that estimate token-specific steering coefficients from the activations themselves and are trained to imitate prompt-based interventions. Experiments on three steering benchmarks across multiple language models show that PSR models outperform existing activation steering methods, especially when controlling for high-coherence completions, and also compare favorably to prompting on AxBench and persona steering.

Abstract

Language Model Steering

Explore similar work

Steer Like the LLM: Activation Steering that Mimics Prompting

From Words to Widgets for Controllable LLM Generation

Can Activation Steering Capture Multidimensional Authorship Style?