Cascaded Sparse Autoencoders Learn Multi-Level Visual Concepts in Multimodal LLMs
Multimodal Large Language Models (MLLMs) have demonstrated strong performance on vision-language tasks, yet their internal visual representations remain difficult to interpret. Sparse Autoencoders (SAEs) provide a scalable way to decompose dense model activations into sparse, interpretable features. However, existing SAE architectures primarily recover flat feature dictionaries and are less suited for explicit multi-level concept organization. In this paper, we introduce cascaded sparse autoencoders (CSAEs) for learning hierarchical visual concepts in MLLMs. Rather than nesting or stacking SAE sparse activation codes, CSAEs train a second-level SAE directly on the decoder weights of the first-level SAE, treating learned low-level feature directions as inputs for higher-level abstraction. This design enables CSAEs to learn "concepts of concepts" while avoiding drawbacks from the shared-prefix coupling of nesting, Matryoshka-style hierarchies and the bottlenecks of naively stacked SAEs. Experiments across Qwen3-VL, Gemma-3, and LLaVA on multiple visual datasets show that CSAEs improve interpretability in terms of hierarchical concept coherence over state-of-the-art SAE baselines. Results on concept steering further demonstrate that the learned concept groups support effective group-level interventions in MLLM outputs.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
How LLMs Learn: Tracing Internal Representations with Sparse Autoencoders
Large Language Models (LLMs) demonstrate remarkable multilingual capabilities and broad knowledge. However, the internal mechanisms underlying the development of these capabilities remain poorly understood. To investigat…
TIDE : Temporal-Aware Sparse Autoencoders for Interpretable Diffusion Transformers in Image Generation
Diffusion Transformers (DiTs) are a powerful yet underexplored class of generative models compared to U-Net-based diffusion models. To bridge this gap, we introduce TIDE (Temporal-aware Sparse Autoencoders for Interpreta…
DenoisingImage GenerationStyle TransferLearning Multi-Level Features with Matryoshka Sparse Autoencoders
Sparse autoencoders (SAEs) have emerged as a powerful tool for interpreting neural networks by extracting the concepts represented in their activations. However, choosing the size of the SAE dictionary (i.e. number of le…
Memory-efficient High-resolution OCT Volume Synthesis with Cascaded Amortized Latent Diffusion Models
Optical coherence tomography (OCT) image analysis plays an important role in the field of ophthalmology. Current successful analysis models rely on available large datasets, which can be challenging to be obtained for ce…
Temporal Sparse Autoencoders: Leveraging the Sequential Nature of Language for Interpretability
Translating the internal representations and computations of models into concepts that humans can understand is a key goal of interpretability. While recent dictionary learning methods such as Sparse Autoencoders (SAEs) …