相关论文: ConceptViz: A Visual Analytics Approach for Explor…

Sparse Autoencoders (SAEs) have been successfully used to probe Large Language Models (LLMs) and extract interpretable concepts from their internal representations. These concepts are linear combinations of neuron activations that…

Developing human understandable interpretation of large language models (LLMs) becomes increasingly critical for their deployment in essential domains. Mechanistic interpretability seeks to mitigate the issues through extracts…

Unsupervised approaches to large language model (LLM) interpretability, such as sparse autoencoders (SAEs), offer a way to decode LLM activations into interpretable and, ideally, controllable concepts. On the one hand, these approaches…

With the integration of image modality, the semantic space of multimodal large language models (MLLMs) is more complex than text-only models, making their interpretability more challenging and their alignment less stable, particularly…

Large Language Models (LLMs) demonstrate exceptional capabilities in factual question answering, yet they sometimes provide incorrect responses. To address this issue, knowledge editing techniques have emerged as effective methods for…

Current multimodal large language models (MLLMs), while effective in natural image understanding, struggle with visualization understanding due to their inability to decode the data-to-visual mapping and extract structured information. To…

Large language models (LLMs) have shown promising results for software engineering applications, but still struggle with code reasoning tasks such as vulnerability detection (VD). We introduce ConceptCoder, a fine-tuning method that…

Concept-based explanations work by mapping complex model computations to human-understandable concepts. Evaluating such explanations is very difficult, as it includes not only the quality of the induced space of possible concepts but also…

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 rapidly increasing in capability, but they still struggle with highly specialized programming tasks such as scientific visualization. We present an LLM assistant, ChatVis, that aids the LLM to generate…

Unlearning methods for vision-language models (VLMs) have primarily adapted techniques from large language models (LLMs), relying on weight updates that demand extensive annotated forget sets. Moreover, these methods perform unlearning at a…

The ability of large language models (LLMs) to interpret visual representations of data is crucial for advancing their application in data analysis and decision-making processes. This paper presents a novel synthetic dataset designed to…

Understanding the decision-making processes of neural networks is a central goal of mechanistic interpretability. In the context of Large Language Models (LLMs), this involves uncovering the underlying mechanisms and identifying the roles…

The evaluation of large language models (LLMs) relies heavily on standardized benchmarks. These benchmarks provide useful aggregated metrics for a given capability, but those aggregated metrics can obscure (i) particular sub-areas where the…

Designers often create visualizations to achieve specific high-level analytical or communication goals. These goals require people to extract complex and interconnected data patterns. Prior perceptual studies of visualization effectiveness…

Large Multimodal Models (LMMs), or Vision-Language Models (VLMs), have shown impressive capabilities in a wide range of visual tasks. However, they often struggle with fine-grained visual reasoning, failing to identify domain-specific…

Recent growth in the popularity of large language models has led to their increased usage for summarizing, predicting, and generating text, making it vital to help researchers and engineers understand how and why they work. We present…

Large-scale vision-language models (VLMs), such as CLIP, have achieved remarkable success in zero-shot learning (ZSL) by leveraging large-scale visual-text pair datasets. However, these methods often lack interpretability, as they compute…

Sparse autoencoders (SAEs) have emerged as a powerful tool for uncovering interpretable features in large language models (LLMs) through the sparse directions they learn. However, the sheer number of extracted directions makes comprehensive…

Translating natural language to visualization (NL2VIS) has shown great promise for visual data analysis, but it remains a challenging task that requires multiple low-level implementations, such as natural language processing and…