相关论文: Vector Arithmetic in Concept and Token Subspaces

Large language models (LLMs) have demonstrated remarkable capabilities, but they still frequently produce hallucinations. These hallucinations are difficult to detect in reasoning-intensive tasks, where the content appears coherent but…

We propose a novel class of language models, Latent Thought Models (LTMs), which incorporate explicit latent thought vectors that follow an explicit prior model in latent space. These latent thought vectors guide the autoregressive…

While Transformers have rapidly gained popularity in various computer vision applications, post-hoc explanations of their internal mechanisms remain largely unexplored. Vision Transformers extract visual information by representing image…

Latent actions serve as an intermediate representation that enables consistent modeling of vision-language-action (VLA) models across heterogeneous datasets. However, approaches to supervising VLAs with latent actions are fragmented and…

Logit Lens has been proposed for visualizing tokens that contribute most to LLM answers. Recently, Logit Lens was also shown to be applicable in autoregressive Vision-Language Models (VLMs), where it illustrates the conceptual content of…

Existing Multimodal Large Language Models (MLLMs) process a large number of visual tokens, leading to significant computational costs and inefficiency. Instruction-related visual token compression demonstrates strong task relevance, which…

Handwritten mathematical expressions (HMEs) contain ambiguities in their interpretations, even for humans sometimes. Several math symbols are very similar in the writing style, such as dot and comma or 0, O, and o, which is a challenge for…

Multimodal language models (MLMs) still face challenges in fundamental visual perception tasks where specialized models excel. Tasks requiring reasoning about 3D structures benefit from depth estimation, and reasoning about 2D object…

The generation of factually incorrect objects, commonly known as object hallucination, remains a persistent challenge in Large Vision-Language Models (LVLMs). Current approaches to address this issue - ranging from expensive data-driven…

Recent empirical evidence shows that LLM representations encode human-interpretable concepts. Nevertheless, the mechanisms by which these representations emerge remain largely unexplored. To shed further light on this, we introduce a novel…

Production LLM systems often rely on separate models for safety and other classification-heavy steps, increasing latency, VRAM footprint, and operational complexity. We instead reuse computation already paid for by the serving LLM: we train…

While Transformer self-attention offers strong parallelism, the Key-Value (KV) cache grows linearly with sequence length and becomes a bottleneck for inference efficiency. Multi-head latent attention was recently developed to compress the…

Large Language Models (LLMs) apply uniform computation to all tokens, despite language exhibiting highly non-uniform information density. This token-uniform regime wastes capacity on locally predictable spans while under-allocating…

Vision-Language Models (VLMs) have become essential backbones of modern multimodal intelligence, yet their outputs remain prone to hallucination-plausible text misaligned with visual inputs. Existing alignment approaches often rely on…

Large Vision-Language Models (LVLMs) can reason effectively over both textual and visual inputs, but they tend to hallucinate syntactically coherent yet visually ungrounded contents. In this paper, we investigate the internal dynamics of…

Large Vision Language Models show impressive performance across image and video understanding tasks, yet their computational cost grows rapidly with the number of visual tokens. Existing token pruning methods mitigate this issue through…

The opaque nature of Large Language Models (LLMs) has led to significant research efforts aimed at enhancing their interpretability, primarily through post-hoc methods. More recent in-hoc approaches, such as Concept Bottleneck Models…

Large Vision-Language Models (LVLMs) integrate image encoders with Large Language Models (LLMs) to process multi-modal inputs and perform complex visual tasks. However, they often generate hallucinations by describing non-existent objects…

Hallucination detection is critical for ensuring the reliability of large language models (LLMs) in context-based generation. Prior work has explored intrinsic signals available during generation, among which attention offers a direct view…

Recently, multimodal large language models (MLLMs) have emerged as a key approach in achieving artificial general intelligence. In particular, vision-language MLLMs have been developed to generate not only text but also visual outputs from…