Investigating Linguistic Steering: An Analysis of Adjectival Effects Across Large Language Model Architectures

• URL: http://arxiv.org/abs/2606.20572v1
• Date: Tue, 28 Apr 2026 12:14:30 GMT
• Title: Investigating Linguistic Steering: An Analysis of Adjectival Effects Across Large Language Model Architectures
• Abstract summary: We use Shapley values to quantify the steering effect of individual adjectives on model performance.<n>We find that a small subset of adjectives act as disproportionately powerful "levers," yet their effects are not universal.<n>Cross-model analysis reveals a "family effect," while architecturally distinct models react in a largely uncorrelated manner.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Achieving reliable control of Large Language Models (LLMs) requires a precise, scalable understanding of how they interpret linguistic cues. We introduce a rigorous framework using Shapley values to quantify the steering effect of individual adjectives on model performance, moving beyond anecdotal heuristics to principled attribution. Applying this method to 100 adjectives across a diverse suite of models (including o3, gpt-4o-mini, phi-3, llama-3-70b, and deepseek-r1) on the MMLU benchmark, we uncover several critical findings for AI alignment. First, we find that a small subset of adjectives act as disproportionately powerful "levers," yet their effects are not universal. Cross-model analysis reveals a "family effect": models of a shared lineage exhibit correlated sensitivity profiles, while architecturally distinct models react in a largely uncorrelated manner, challenging the notion of a one-size-fits-all prompting strategy. Second, focused follow-up studies demonstrate that the steering direction of these powerful adjectives is not intrinsic but is highly contingent on their syntactic role and position within the prompt. For larger models like gpt-4o-mini, we provide the first quantitative evidence of strong, non-additive interaction effects where adjectives can synergistically amplify, antagonistically dampen, or even reverse each other's impact. In contrast, smaller models like phi-3 exhibit a more literal and less compositional response. These results suggest that as models scale, their interpretation of prompts becomes more sophisticated but also less predictable, posing a significant challenge for robustly steering model behavior and highlighting the need for compositional and model-specific alignment techniques.

Related papers

• The Impact of Steering Large Language Models with Persona Vectors in Educational Applications [2.2917707112773598]
  Activation-based steering can personalize large language models at inference time.<n>This study systematically examines the effects of activation-steered persona traits in educational generation and scoring.
  arXiv  Detail & Related papers  (2026-04-08T13:55:32Z)
• Preference learning in shades of gray: Interpretable and bias-aware reward modeling for human preferences [8.680081568962999]
  This study investigates the limits of current approaches to learning human preferences in language models.<n>We propose a feature-augmented framework to better capture the multidimensional nature of human judgment.
  arXiv  Detail & Related papers  (2026-04-01T18:26:16Z)
• UniG2U-Bench: Do Unified Models Advance Multimodal Understanding? [50.92401586025528]
  Unified multimodal models have recently demonstrated strong generative capabilities, yet whether and when generation improves understanding remains unclear.<n>We introduce UniG2U-Bench, a comprehensive benchmark categorizing generation-to-understanding (G2U) evaluation into 7 regimes and 30 subtasks.
  arXiv  Detail & Related papers  (2026-03-03T18:36:16Z)
• Plug-and-Play Clarifier: A Zero-Shot Multimodal Framework for Egocentric Intent Disambiguation [60.63465682731118]
  The performance of egocentric AI agents is fundamentally limited by multimodal intent ambiguity.<n>We introduce the Plug-and-Play Clarifier, a zero-shot and modular framework that decomposes the problem into discrete, solvable sub-tasks.<n>Our framework improves the intent clarification performance of small language models by approximately 30%, making them competitive with significantly larger counterparts.
  arXiv  Detail & Related papers  (2025-11-12T04:28:14Z)
• Stands to Reason: Investigating the Effect of Reasoning on Idiomaticity Detection [2.8330244018167945]
  We examine how reasoning capabilities in Large Language Models affect idiomaticity detection performance.<n>We find the effect of reasoning to be smaller and more varied than expected.<n>For smaller models, producing chain-of-thought (CoT) reasoning increases performance from Math-tuned intermediate models, but not to the levels of the base models.
  arXiv  Detail & Related papers  (2025-08-18T21:17:09Z)
• Reasoning Capabilities of Large Language Models on Dynamic Tasks [0.017476232824732776]
  Large language models excel on static benchmarks, but their ability as self-learning agents in dynamic environments remains unclear.<n>We evaluate three prompting strategies: self-reflection, mutation, and planning across dynamic tasks with open-source models.<n>We find that larger models generally outperform smaller ones, but that strategic prompting can close this performance gap.
  arXiv  Detail & Related papers  (2025-05-15T17:53:47Z)
• Human-Object Interaction Detection Collaborated with Large Relation-driven Diffusion Models [65.82564074712836]
  We introduce DIFfusionHOI, a new HOI detector shedding light on text-to-image diffusion models. We first devise an inversion-based strategy to learn the expression of relation patterns between humans and objects in embedding space. These learned relation embeddings then serve as textual prompts, to steer diffusion models generate images that depict specific interactions.
  arXiv  Detail & Related papers  (2024-10-26T12:00:33Z)
• Why Larger Language Models Do In-context Learning Differently? [12.554356517949785]
  Large language models (LLM) have emerged as a powerful tool for AI, with the key ability of in-context learning (ICL) One recent mysterious observation is that models of different scales may have different ICL behaviors.
  arXiv  Detail & Related papers  (2024-05-30T01:11:35Z)
• Training Trajectories of Language Models Across Scales [99.38721327771208]
  Scaling up language models has led to unprecedented performance gains. How do language models of different sizes learn during pre-training? Why do larger language models demonstrate more desirable behaviors?
  arXiv  Detail & Related papers  (2022-12-19T19:16:29Z)
• Fairness Increases Adversarial Vulnerability [50.90773979394264]
  This paper shows the existence of a dichotomy between fairness and robustness, and analyzes when achieving fairness decreases the model robustness to adversarial samples. Experiments on non-linear models and different architectures validate the theoretical findings in multiple vision domains. The paper proposes a simple, yet effective, solution to construct models achieving good tradeoffs between fairness and robustness.
  arXiv  Detail & Related papers  (2022-11-21T19:55:35Z)
• AES Systems Are Both Overstable And Oversensitive: Explaining Why And Proposing Defenses [66.49753193098356]
  We investigate the reason behind the surprising adversarial brittleness of scoring models. Our results indicate that autoscoring models, despite getting trained as "end-to-end" models, behave like bag-of-words models. We propose detection-based protection models that can detect oversensitivity and overstability causing samples with high accuracies.
  arXiv  Detail & Related papers  (2021-09-24T03:49:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.