mlsanityData to 7 Oct 2026

Style Vectors for Steering Generative Large Language Model

arXiv 2402.01618 ↗First posted 2 Feb 2024cs.CLcited by 56

Not on the map. Why?

Abstract (arXiv)

This research explores strategies for steering the output of large language models (LLMs) towards specific styles, such as sentiment, emotion, or writing style, by adding style vectors to the activations of hidden layers during text generation. We show that style vectors can be simply computed from recorded layer activations for input texts in a specific style in contrast to more complex training-based approaches. Through a series of experiments, we demonstrate the effectiveness of activation engineering using such style vectors to influence the style of generated text in a nuanced and parameterisable way, distinguishing it from prompt engineering. The presented research constitutes a significant step towards developing more adaptive and effective AI-empowered interactive systems.

Key facts. Cited by 56 papers in the map (data through 7 Oct 2026). Builds on 22 of its 101 references that are in the map. First posted on arXiv 2 Feb 2024.

Builds on

22 of 101 references.

References · 22

  1. 2017Attention Is All You Needcited by 59,932
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2020Language Models are Few-Shot Learnerscited by 32,199
  4. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  5. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  6. 2019Parameter-Efficient Transfer Learning for NLPcited by 4,195
  7. 2021Prefix-Tuning: Optimizing Continuous Prompts for Generationcited by 3,878
  8. 2021The Power of Scale for Parameter-Efficient Prompt Tuningcited by 3,769
  9. 2023Steering Language Models With Activation Engineeringcited by 824
  10. 2019CTRL: A Conditional Transformer Language Model for Controllable Generationcited by 754
  11. 2022Safe Latent Diffusion: Mitigating Inappropriate Degeneration in Diffusion Modelscited by 501
  12. 2020GoEmotions: A Dataset of Fine-Grained Emotionscited by 479
10 more
  1. 2021AI Chains: Transparent and Controllable Human-AI Interaction by Chaining Large Language Model Promptscited by 265
  2. 2017Style Transfer from Non-Parallel Text by Cross-Alignmentcited by 246
  3. 2021Large Pre-trained Language Models Contain Human-like Biases of What is Right and Wrong to Docited by 178
  4. 2020Deep Learning for Text Style Transfer: A Surveycited by 161
  5. 2017Shakespearizing Modern Language Using Copy-Enriched Sequence-to-Sequence Modelscited by 57
  6. 2020Hooks in the Headline: Learning to Generate Headlines with Controlled Stylescited by 46
  7. 2019Structuring Latent Spaces for Stylized Response Generationcited by 25
  8. 2022The Stable Artist: Steering Semantics in Diffusion Latent Spacecited by 23
  9. 2020StyleDGPT: Stylized Response Generation with Pre-trained Language Modelscited by 10
  10. 2022Stylized Knowledge-Grounded Dialogue Generation via Disentangled Template Rewritingcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 5; 2025: 25; 2026: 26.

56 papers.

Sort:

By area:

  1. 2024Programming Refusal with Conditional Activation Steeringcited by 191
  2. 2025MemOS: A Memory OS for AI Systemcited by 129
  3. 2025Angular Steering: Behavior Control via Rotation in Activation Spacecited by 45
  4. 2025Understanding (Un)Reliability of Steering Vectors in Language Modelscited by 39
  5. 2024Beyond Single Concept Vector: Modeling Concept Subspace in LLMs with Gaussian Distributioncited by 24
  6. 2025Beyond Prompt Engineering: Robust Behavior Control in LLMs via Steering Target Atomscited by 23
  7. 2024LiSeCo: Linear Semantic Control for Language Generationcited by 20
  8. 2025The Rogue Scalpel: Activation Steering Compromises LLM Safetycited by 18
  9. 2025ThinkEdit: Interpretable Weight Editing to Mitigate Overly Short Thinking in Reasoning Modelscited by 17
  10. 2025Differentially Private Steering for Large Language Model Alignmentcited by 15
  11. 2026Valence-Arousal Subspace in LLMs: Circular Emotion Geometry and Multi-Behavioral Controlcited by 12
  12. 2025Activation Steering with a Feedback Controllercited by 11
38 more
  1. 2025Multi-objective Large Language Model Alignment with Hierarchical Expertscited by 10
  2. 2024Words in Motion: Extracting Interpretable Control Vectors for Motion Transformerscited by 9
  3. 2026Why Steering Works: Toward a Unified View of Language Model Parameter Dynamicscited by 8
  4. 2025Effectively Steer LLM To Follow Preference via Building Confident Directionscited by 7
  5. 2025Mitigating Hallucination in VideoLLMs via Temporal-Aware Activation Engineeringcited by 7
  6. 2025Activation Steering for Masked Diffusion Language Modelscited by 7
  7. 2026Steer2Adapt: Dynamically Composing Steering Vectors Elicits Efficient Adaptation of LLMscited by 7
  8. 2025Silence is Golden: Mitigating Hallucinations in Large Audio-Language Models via Layer-Weighted Vector Steeringcited by 6
  9. 2026Efficient and accurate steering of Large Language Models through attention-guided feature learningcited by 6
  10. 2026Fine-Grained Activation Steering: Steering Less, Achieving Morecited by 6
  11. 2025What's Producible May Not Be Reachable: Measuring the Steerability of Generative Modelscited by 5
  12. 2026EmoShift: Lightweight Activation Steering for Enhanced Emotion-Aware Speech Synthesiscited by 5
  13. 2025Bottlenecked Transformers: Periodic KV Cache Consolidation for Generalised Reasoningcited by 3
  14. 2025We Think, Therefore We Align LLMs to Helpful, Harmless and Honest Before They Go Wrongcited by 3
  15. 2025Machine Unlearning Meets Adversarial Robustness via Constrained Interventions on LLMscited by 3
  16. 2026The Effectiveness of Style Vectors for Steering Large Language Models: A Human Evaluationcited by 3
  17. 2025PersRM-R1: Enhance Personalized Reward Modeling with Reinforcement Learningcited by 2
  18. 2024AI Assistants for Spaceflight Procedures: Combining Generative Pre-Trained Transformer and Retrieval-Augmented Generation on Knowledge Graphs With Augmented Reality Cuescited by 1
  19. 2025Fusion Steering: Prompt-Specific Activation Controlcited by 1
  20. 2025Command-V: Pasting LLM Behaviors via Activation Profilescited by 1
  21. 2025VISOR++: Universal Visual Inputs based Steering for Large Vision Language Modelscited by 1
  22. 2025Confidence is Not Competencecited by 1
  23. 2026Circuit Fingerprints: How Answer Tokens Encode Their Geometrical Pathcited by 1
  24. 2026Control Reinforcement Learning: Interpretable Token-Level Steering of LLMs via Sparse Autoencoder Featurescited by 1
  25. 2026Quantitative Introspection in Language Models: Tracking Emotive States Across Conversationcited by 1
  26. 2026Mechanistic Analysis of Alignment Algorithms in Language Modelscited by 1
  27. 2026Steering Where to Listen: Instruction-Based Activation Steering Redirects Temporal Attention in Large Audio-Language Modelscited by 1
  28. 2026DecodeShare: Tracing the Shared Subspace of LLM Decode-Time Decisionscited by 1
  29. 2026Read-Best Is Not Steer-Best: A Probing--Steering Layer Dissociation in Omni-Modal Large Language Modelscited by 1
  30. 2025Co-creation for Sign Language Processing and Machine Translationcited by 0
  31. 2025Mind the Quote: Enabling Quotation-Aware Dialogue in LLMs via Plug-and-Play Modulescited by 0
  32. 2025VISOR: Visual Input-based Steering for Output Redirection in Vision-Language Modelscited by 0
  33. 2026Evaluating Prompt Engineering Strategies for Sentiment Control in AI-Generated Textscited by 0
  34. 2026Facet-Level Persona Control by Trait-Activated Routing with Contrastive SAE for Role-Playing LLMscited by 0
  35. 2026The Master Key Hypothesis: Unlocking Cross-Model Capability Transfer via Linear Subspace Alignmentcited by 0
  36. 2026HAS: Highlight-guided Attention Steering for Multimodal LLM Video Summarizationcited by 0
  37. 2026Probabilistic Concept-Aware Steering for Trustworthy LLM Inferencecited by 0
  38. 2026Latent Fact-Checking: Detecting Misinformation through Activation Engineeringcited by 0

50 most cited of 56 All 56 (JSON)