LF-Steering: Latent Feature Activation Steering for Enhancing Semantic Consistency in Large Language Models

• URL: http://arxiv.org/abs/2501.11036v2
• Date: Wed, 22 Jan 2025 13:53:59 GMT
• Title: LF-Steering: Latent Feature Activation Steering for Enhancing Semantic Consistency in Large Language Models
• Authors: Jingyuan Yang, Rongjun Li, Weixuan Wang, Ziyu Zhou, Zhiyong Feng, Wei Peng,
• Abstract summary: Large Language Models (LLMs) often generate inconsistent responses when prompted with semantically equivalent paraphrased inputs.<n>We propose LF-Steering, a novel activation steering approach to precisely identify latent feature representations responsible for semantic inconsistency.<n>Our method maps the hidden states of the relevant transformer layer into a sparsely activated, high-dimensional feature space based on a sparse autoencoder.
• Score: 16.37602070339033
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large Language Models (LLMs) often generate inconsistent responses when prompted with semantically equivalent paraphrased inputs. Recently, activation steering, a technique that modulates LLMs' behaviours by adjusting their latent representations during inference time, has been explored to improve the semantic consistency of LLMs. However, these methods typically operate at the model component level, such as layer hidden states or attention head outputs. They face a challenge due to the ``polysemanticity issue'', where the model components of LLMs typically encode multiple entangled features, making precise steering difficult. To address this challenge, we drill down to feature-level representations and propose LF-Steering, a novel activation steering approach to precisely identify latent feature representations responsible for semantic inconsistency. More specifically, our method maps the hidden states of the relevant transformer layer into a sparsely activated, high-dimensional feature space based on a sparse autoencoder (SAE), ensuring model steering based on decoupled feature representations with minimal interference. Comprehensive experiments on NLU and NLG datasets demonstrate the effectiveness of our method in enhancing semantic consistency, resulting in significant performance gains for various NLU and NLG tasks.

Related papers

• Enhancing LLM Robustness to Perturbed Instructions: An Empirical Study [8.827173113748701]
  We study character- and word-level edits of task-specific instructions, which substantially degrade downstream performance. We find that, on average, self-denoising achieves substantially higher performance gains than alternative strategies.
  arXiv  Detail & Related papers  (2025-04-03T16:17:56Z)
• Towards LLM Guardrails via Sparse Representation Steering [11.710399901426873]
  Large Language Models (LLMs) have demonstrated remarkable performance in natural language generation tasks. We propose a sparse encoding-based representation engineering method, named SRE, which decomposes polysemantic activations into a structured, monosemantic feature space. By leveraging sparse autoencoding, our approach isolates and adjusts only task-specific sparse feature dimensions, enabling precise and interpretable steering of model behavior.
  arXiv  Detail & Related papers  (2025-03-21T04:50:25Z)
• Refining Sentence Embedding Model through Ranking Sentences Generation with Large Language Models [60.00178316095646]
  Sentence embedding is essential for many NLP tasks, with contrastive learning methods achieving strong performance using datasets like NLI. Recent studies leverage large language models (LLMs) to generate sentence pairs, reducing annotation dependency. We propose a method for controlling the generation direction of LLMs in the latent space. Unlike unconstrained generation, the controlled approach ensures meaningful semantic divergence. Experiments on multiple benchmarks demonstrate that our method achieves new SOTA performance with a modest cost in ranking sentence synthesis.
  arXiv  Detail & Related papers  (2025-02-19T12:07:53Z)
• Enhancing Semantic Consistency of Large Language Models through Model Editing: An Interpretability-Oriented Approach [28.07366458452159]
  Large Language Models (LLM) generate inconsistent and sometimes contradictory outputs when presented with a prompt that has equivalent semantics but is expressed differently from the original prompt.<n>To achieve semantic consistency of an LLM, one of the key approaches is to finetune the model with prompt-output pairs with semantically equivalent meanings.<n>We propose a more interpretable method (i.e., model editing) to enhance the semantic consistency of LLMs.
  arXiv  Detail & Related papers  (2025-01-19T13:26:15Z)
• Enhancing Multiple Dimensions of Trustworthiness in LLMs via Sparse Activation Control [44.326363467045496]
  Large Language Models (LLMs) have become a critical area of research in Reinforcement Learning from Human Feedback (RLHF) representation engineering offers a new, training-free approach. This technique leverages semantic features to control the representation of LLM's intermediate hidden states. It is difficult to encode various semantic contents, like honesty and safety, into a singular semantic feature.
  arXiv  Detail & Related papers  (2024-11-04T08:36:03Z)
• Semantics-Adaptive Activation Intervention for LLMs via Dynamic Steering Vectors [8.761404991620285]
  Activation intervention has emerged as an effective and economical method to modify the behavior of large language models (LLMs) We propose Semantics-Adaptive Dynamic Intervention (SADI), a novel method that constructs a dynamic steering vector to intervene model activations at inference time. Experimental results show that SADI outperforms established baselines by substantial margins, improving task performance without training.
  arXiv  Detail & Related papers  (2024-10-16T06:58:49Z)
• Aligning Large Language Models with Representation Editing: A Control Perspective [38.71496554018039]
  Fine-tuning large language models (LLMs) to align with human objectives is crucial for real-world applications. Test-time alignment techniques, such as prompting and guided decoding, do not modify the underlying model. We propose aligning LLMs through representation editing.
  arXiv  Detail & Related papers  (2024-06-10T01:21:31Z)
• Modality Prompts for Arbitrary Modality Salient Object Detection [57.610000247519196]
  This paper delves into the task of arbitrary modality salient object detection (AM SOD) It aims to detect salient objects from arbitrary modalities, eg RGB images, RGB-D images, and RGB-D-T images. A novel modality-adaptive Transformer (MAT) will be proposed to investigate two fundamental challenges of AM SOD.
  arXiv  Detail & Related papers  (2024-05-06T11:02:02Z)
• Characterizing Truthfulness in Large Language Model Generations with Local Intrinsic Dimension [63.330262740414646]
  We study how to characterize and predict the truthfulness of texts generated from large language models (LLMs) We suggest investigating internal activations and quantifying LLM's truthfulness using the local intrinsic dimension (LID) of model activations.
  arXiv  Detail & Related papers  (2024-02-28T04:56:21Z)
• Entropy-Regularized Token-Level Policy Optimization for Language Agent Reinforcement [67.1393112206885]
  Large Language Models (LLMs) have shown promise as intelligent agents in interactive decision-making tasks. We introduce Entropy-Regularized Token-level Policy Optimization (ETPO), an entropy-augmented RL method tailored for optimizing LLMs at the token level. We assess the effectiveness of ETPO within a simulated environment that models data science code generation as a series of multi-step interactive tasks.
  arXiv  Detail & Related papers  (2024-02-09T07:45:26Z)
• InferAligner: Inference-Time Alignment for Harmlessness through Cross-Model Guidance [56.184255657175335]
  We develop textbfInferAligner, a novel inference-time alignment method that utilizes cross-model guidance for harmlessness alignment. Experimental results show that our method can be very effectively applied to domain-specific models in finance, medicine, and mathematics. It significantly diminishes the Attack Success Rate (ASR) of both harmful instructions and jailbreak attacks, while maintaining almost unchanged performance in downstream tasks.
  arXiv  Detail & Related papers  (2024-01-20T10:41:03Z)
• Improve Variational Autoencoder for Text Generationwith Discrete Latent Bottleneck [52.08901549360262]
  Variational autoencoders (VAEs) are essential tools in end-to-end representation learning. VAEs tend to ignore latent variables with a strong auto-regressive decoder. We propose a principled approach to enforce an implicit latent feature matching in a more compact latent space.
  arXiv  Detail & Related papers  (2020-04-22T14:41:37Z)

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.