SAEVerbalizer: Generating Explanations for Sparse Autoencoder Features via Representation Verbalization

• URL: http://arxiv.org/abs/2608.13538v2
• Date: Tue, 18 Aug 2026 07:35:07 GMT
• Title: SAEVerbalizer: Generating Explanations for Sparse Autoencoder Features via Representation Verbalization
• Abstract summary: We introduce SAEVerbalizer, a framework that injects SAE decoder directions into a large language model's representations.<n>The resulting verbalizer explains SAE features directly from decoder directions, addressing both limitations.<n>Experiments show that the learned verbalization capability generalizes to unseen features, transfers across separately trained SAE dictionaries, and, with a lightweight adapter, extends to SAE features from different LLMs.
• Score: 45.84399159937069
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Sparse autoencoders (SAEs) are proposed to extract numerous features from large language model (LLM) representations, yet explaining these features still relies primarily on external observation. This reliance leads to superficial explanations inferred from observed model behavior and computational inefficiency from collecting such behavioral evidence at scale. We introduce SAEVerbalizer, a framework that injects SAE decoder directions into an LLM's representations and fine-tunes the LLM's downstream layers to generate natural-language explanations of the injected features. Once trained, the resulting verbalizer explains SAE features directly from decoder directions, addressing both limitations. Our experiments show that the learned verbalization capability generalizes to unseen features, transfers across separately trained SAE dictionaries, and, with a lightweight adapter, extends to SAE features from different LLMs. Intervention experiments show that injecting multiple directions yields an explanation combining their meanings, while reversing individual directions produces corresponding meaning shifts.

Related papers

• Step-Level Sparse Autoencoder for Reasoning Process Interpretation [48.99201531966593]
  Large Language Models (LLMs) have achieved strong complex reasoning capabilities through Chain-of-Thought (CoT) reasoning.<n>We propose step-level sparse autoencoder (SSAE), which serves as an analytical tool to disentangle different aspects of LLMs' reasoning steps into sparse features.<n> Experiments on multiple base models and reasoning tasks show the effectiveness of the extracted features.
  arXiv  Detail & Related papers  (2026-03-03T14:25:02Z)
• SAGE: An Agentic Explainer Framework for Interpreting SAE Features in Language Models [37.102387880457535]
  Large language models (LLMs) have achieved remarkable progress, yet their internal mechanisms remain largely opaque.<n>Sparse autoencoders (SAEs) have emerged as a promising tool for decomposing LLM representations into more interpretable features.<n>We propose SAGE (SAE AGentic Explainer), an agent-based framework that recasts feature interpretation from a passive, single-pass generation task into an explanationdriven process.
  arXiv  Detail & Related papers  (2025-11-25T20:14:29Z)
• ProtSAE: Disentangling and Interpreting Protein Language Models via Semantically-Guided Sparse Autoencoders [30.219733023958188]
  Sparse Autoencoder (SAE) has emerged as a powerful tool for mechanistic interpretability of large language models.<n>We propose a semantically-guided SAE, called ProtSAE.<n>We show that ProtSAE learns more biologically relevant and interpretable hidden features compared to previous methods.
  arXiv  Detail & Related papers  (2025-08-26T11:20:31Z)
• Ensembling Sparse Autoencoders [10.81463830315253]
  Sparse autoencoders (SAEs) are used to decompose neural network activations into human-interpretable features.<n>We propose to ensemble multiple SAEs through naive bagging and boosting.<n>Our empirical results demonstrate that ensembling SAEs can improve the reconstruction of language model activations, diversity of features, and SAE stability.
  arXiv  Detail & Related papers  (2025-05-21T23:31:21Z)
• Sparse Autoencoders Learn Monosemantic Features in Vision-Language Models [50.587868616659826]
  We introduce a comprehensive framework for evaluating monosemanticity at the neuron-level in vision representations.<n>Our experimental results reveal that SAEs trained on Vision-Language Models significantly enhance the monosemanticity of individual neurons.
  arXiv  Detail & Related papers  (2025-04-03T17:58:35Z)
• Interpreting and Steering LLMs with Mutual Information-based Explanations on Sparse Autoencoders [29.356200147371275]
  Large language models (LLMs) excel at handling human queries, but they can occasionally generate flawed or unexpected responses.<n>We propose using a fixed vocabulary set for feature interpretations and designing a mutual information-based objective.<n>We propose two runtime steering strategies that adjust the learned feature activations based on their corresponding explanations.
  arXiv  Detail & Related papers  (2025-02-21T16:36:42Z)
• LatentQA: Teaching LLMs to Decode Activations Into Natural Language [72.87064562349742]
  We introduce LatentQA, the task of answering open-ended questions about model activations in natural language.<n>We propose Latent Interpretation Tuning (LIT), which finetunes a decoder LLM on a dataset of activations and associated question-answer pairs.<n>Our decoder also specifies a differentiable loss that we use to control models, such as debiasing models on stereotyped sentences and controlling the sentiment of generations.
  arXiv  Detail & Related papers  (2024-12-11T18:59:33Z)
• When Backdoors Speak: Understanding LLM Backdoor Attacks Through Model-Generated Explanations [58.27927090394458]
  Large Language Models (LLMs) are known to be vulnerable to backdoor attacks.<n>In this paper, we examine backdoor attacks through the novel lens of natural language explanations.<n>Our results show that backdoored models produce coherent explanations for clean inputs but diverse and logically flawed explanations for poisoned data.
  arXiv  Detail & Related papers  (2024-11-19T18:11:36Z)
• Automatically Interpreting Millions of Features in Large Language Models [1.8035046415192353]
  sparse autoencoders (SAEs) can be used to transform activations into a higher-dimensional latent space.<n>We build an open-source pipeline to generate and evaluate natural language explanations for SAE features.<n>Our large-scale analysis confirms that SAE latents are indeed much more interpretable than neurons.
  arXiv  Detail & Related papers  (2024-10-17T17:56:01Z)
• FaithLM: Towards Faithful Explanations for Large Language Models [60.45183469474916]
  We introduce FaithLM, a model-agnostic framework that evaluates and improves the faithfulness of large language models.<n>We show that FaithLM consistently increases faithfulness and produces explanations more aligned with human rationales than strong self-explanation baselines.
  arXiv  Detail & Related papers  (2024-02-07T09:09:14Z)
• Towards ASR Robust Spoken Language Understanding Through In-Context Learning With Word Confusion Networks [68.79880423713597]
  We introduce a method that utilizes the ASR system's lattice output instead of relying solely on the top hypothesis. Our in-context learning experiments, covering spoken question answering and intent classification, underline the LLM's resilience to noisy speech transcripts.
  arXiv  Detail & Related papers  (2024-01-05T17:58:10Z)

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.