AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Causal-Steer: Disentangled Continuous Style Control without Parallel Corpora

Qingsong Wang, Chang Yao, Jingyuan Chen

ICLR 2026 · Large language models › Language, safety and interpretability · approach: Task vectors and steering

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add contrastive loss to LoRA training (ablation)Causal-Steer · ELI5 conceptual complexity (steer to Complex)Knowledge Difficulty (Complex): 8.24 -> 8.42Improved
add contrastive loss to LoRA training (ablation)Causal-Steer · ELI5 conceptual complexity (steer to Simple)Knowledge Difficulty (Simple, lower better): 3.01 -> 2.77Improved
add differential activation (with vs without LoRA on same input) (ablation)Causal-Steer · ELI5 conceptual complexity (steer to Complex)Knowledge Difficulty (Complex): 7.90 -> 8.42Improved
average activations over generated tokens instead of last token (ablation)Causal-Steer · ELI5 conceptual complexity (steer to Complex)Knowledge Difficulty (Complex): 4.26 -> 8.42Improved
average activations over generated tokens instead of last token (ablation)Causal-Steer · ELI5 conceptual complexity (steer to Simple)Knowledge Difficulty (Simple, lower better): 4.73 -> 2.77Improved
add PCA denoising before aggregation (ablation)Causal-Steer · ELI5 conceptual complexity (steer to Complex)Knowledge Difficulty (Complex): 8.18 -> 8.42Improved
add PCA denoising before aggregation (ablation)Causal-Steer · ELI5 conceptual complexity (steer to Simple)Knowledge Difficulty (Simple, lower better): 2.81 -> 2.77Improved
replace arithmetic mean with geometric median (ablation)Causal-Steer · ELI5 conceptual complexity (steer to Complex)Knowledge Difficulty (Complex): 8.24 -> 8.42Improved
replace arithmetic mean with geometric median (ablation)Causal-Steer · ELI5 conceptual complexity (steer to Simple)Knowledge Difficulty (Simple, lower better): 2.82 -> 2.77No clear effect
train on single-sided Simple dataset only (CSsingle) instead of both stylesCausal-Steer · ELI5 conceptual complexity (steer to Simple)Knowledge Difficulty (Simple, lower better): 2.80 vs 2.77No clear effect
Causal-Steer vs CLMI weight interpolation on formalityCausal-Steer · ELI5 formality (steer to Formal)Formality score (Formal): 7.58 -> 8.61Improved
Causal-Steer vs GCAV-Output on detoxificationCausal-Steer · RealToxicityPrompts (toxic prompts)toxicity (toxic prompts): 0.0879 -> 0.0609Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy