Related papers: Steered Generation via Gradient-Based Optimization…

Understanding the multilingual mechanisms of large language models (LLMs) provides insight into how they process different languages, yet this remains challenging. Existing studies often focus on individual neurons, but their polysemantic…

Semantic occupancy has emerged as a powerful representation in world models for its ability to capture rich spatial semantics. However, most existing occupancy world models rely on static and fixed embeddings or grids, which inherently…

Sparse autoencoders provide a promising unsupervised approach for extracting interpretable features from a language model by reconstructing activations from a sparse bottleneck layer. Since language models learn many concepts, autoencoders…

Structuring the latent space in probabilistic deep generative models, e.g., variational autoencoders (VAEs), is important to yield more expressive models and interpretable representations, and to avoid overfitting. One way to achieve this…

Sparse autoencoders (SAEs) have become a standard tool for mechanistic interpretability in autoregressive large language models (LLMs), enabling researchers to extract sparse, human-interpretable features and intervene on model behavior.…

Sparse autoencoders (SAEs) are a promising unsupervised approach for identifying causally relevant and interpretable linear features in a language model's (LM) activations. To be useful for downstream tasks, SAEs need to decompose LM…

Watermarking LLM-generated text is critical for content attribution and misinformation prevention. However, existing methods compromise text quality, require white-box model access and logit manipulation. These limitations exclude API-based…

Large Language Models (LLMs) have transformed natural language processing, yet their internal mechanisms remain largely opaque. Recently, mechanistic interpretability has attracted significant attention from the research community as a…

Large language models (LLMs) are now ubiquitous in user-facing applications, yet they still generate undesirable toxic outputs, including profanity, vulgarity, and derogatory remarks. Although numerous detoxification methods exist, most…

Discrete diffusion language models (DLMs) generate text by iteratively denoising all positions in parallel, offering an alternative to autoregressive models. Controlled generation methods for DLMs, imported from autoregressive models, apply…

The success of Deep Generative Models at high-resolution image generation has led to their extensive utilization for style editing of real images. Most existing methods work on the principle of inverting real images onto their latent space,…

We develop the sparse VAE for unsupervised representation learning on high-dimensional data. The sparse VAE learns a set of latent factors (representations) which summarize the associations in the observed data features. The underlying…

Despite the state-of-the-art performance of Large Language Models (LLMs), these models often suffer from hallucinations, which can undermine their performance in critical applications. In this work, we propose SAFE, a novel method for…

Large language models (LLMs) can store a significant amount of factual knowledge in their parameters. However, their parametric knowledge may conflict with the information provided in the context -- this phenomenon, known as…

Large language models (LLMs) augmented with retrieval exhibit robust performance and extensive versatility by incorporating external contexts. However, the input length grows linearly in the number of retrieved documents, causing a dramatic…

Large Language Models (LLMs) trained for average correctness often exhibit mode collapse, producing narrow decision behaviors on tasks where multiple responses may be reasonable. This limitation is particularly problematic in ordinal…

Sparse autoencoders (SAEs) are designed to extract interpretable features from language models by enforcing a sparsity constraint. Ideally, training an SAE would yield latents that are both sparse and semantically meaningful. However, many…

Large Language Models (LLMs) demonstrate the ability to solve reasoning and mathematical problems using the Chain-of-Thought (CoT) technique. Expanding CoT length, as seen in models such as DeepSeek-R1, significantly enhances this reasoning…

Video reasoning segmentation requires localizing objects across video frames from natural language expressions, often involving spatial reasoning and implicit references. Recent approaches leverage frozen large vision-language models…

Sparse autoencoders (SAEs) model the activations of a neural network as linear combinations of sparsely occurring directions of variation (latents). The ability of SAEs to reconstruct activations follows scaling laws w.r.t. the number of…