AI Paper Atlas

Research map / Large language models / Alignment and preferences

SDA: Steering-Driven Distribution Alignment for Open LLMs Without Fine-Tuning

Wei Xia, Zhi-Hong Deng

AAAI 2026 · Large language models › Alignment and preferences · approach: Steering and correcting LLM behavior · 1 citations (OpenAlex)

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 steering-based logit realignment to the base model (ablation)Llama2-7B-Chat · Helpfulness (5 datasets)win rate (%): 57.9%Improved
add steering-based logit realignment to the base model (ablation)Llama2-7B-Chat · Honesty (TruthfulQA)win rate (%): 25.9%Improved
add steering-based logit realignment to the base model (ablation)Llama2-7B-Chat · Harmlessnesswin rate (%): 5.3%Improved
add divergence-aware temperature scaling on top of steering (ablation)SDA (Llama2-7B-Chat) · Helpfulnesswin rate (%): 17.9%Improved
add divergence-aware temperature scaling on top of steering (ablation)SDA (Llama2-7B-Chat) · Honestywin rate (%): 5.7%Improved
add divergence-aware temperature scaling on top of steering (ablation)SDA (Llama2-7B-Chat) · Harmlessnesswin rate (%): -2.1%Hurt
apply SDA instead of the trained Aligner-7BEight open LLMs · 3H, helpfulnesswin rate vs Aligner (%): 71.7%Improved
apply SDA instead of the trained Aligner-7BEight open LLMs · Honesty (TruthfulQA)win rate vs Aligner (%): 30.9%Improved
apply SDA instead of the trained Aligner-7BOpen LLMs · Harmlessnesswin rate vs Aligner (%): lags behind Aligner-7B in harmlessness by only 1%No clear effect
apply SDA to the base modelVicuna-13B-V1.5 · E-Dialogue (Empathy)win rate vs Base (%): 64.5% -> 91.6% (full vs refined)Improved
apply SDA to the base modelDeepSeek-R1-Distill-Qwen-14B · HarmfulQA (Harmless)win rate vs Base (%): 21.1% -> -0.7%Mixed
apply SDA to the base modelEight open LLMs · 3H averagewin rate vs Base (%): 64.4% helpfulness, 30% honesty, 11.5% harmlessnessImproved

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