Related papers: Beyond Sequential Distance: Inter-Modal Distance I…

Rotary Positional Embedding (RoPE) is a key component of context scaling in Large Language Models (LLMs). While various methods have been proposed to adapt RoPE to longer contexts, their guiding principles generally fall into two…

While MLLMs perform well on perceptual tasks, they lack precise multimodal alignment, limiting performance. To address this challenge, we propose Vision Dynamic Embedding-Guided Pretraining (VDEP), a hybrid autoregressive training paradigm…

Existing methods in Multimodal Knowledge Editing (MKE) have advanced the ability to correct outdated or inaccurate knowledge in Multimodal Large Language Models (MLLMs). However, they exhibit a critical limitation: while effectively…

Image-text retrieval has developed rapidly in recent years. However, it is still a challenge in remote sensing due to visual-semantic imbalance, which leads to incorrect matching of non-semantic visual and textual features. To solve this…

The excessive use of visual tokens in existing Multimoal Large Language Models (MLLMs) often exhibits obvious redundancy and brings in prohibitively expensive computation. To gain insights into this problem, we first conduct extensive…

Positional encodings are essential to transformer-based generative models, yet their behavior in multimodal and attention-sharing settings is not fully understood. In this work, we present a principled analysis of Rotary Positional…

Since self-attention layers in Transformers are permutation invariant by design, positional encodings must be explicitly incorporated to enable spatial understanding. However, fixed-size lookup tables used in traditional learnable position…

We introduce DiMPLe (Disentangled Multi-Modal Prompt Learning), a novel approach to disentangle invariant and spurious features across vision and language modalities in multi-modal learning. Spurious correlations in visual data often hinder…

Vision transformers have demonstrated significant advantages in computer vision tasks due to their ability to capture long-range dependencies and contextual relationships through self-attention. However, existing position encoding…

Disentanglement learning is central to understanding and reusing learned representations in variational autoencoders (VAEs). Although equivariance has been explored in this context, effectively exploiting it for disentanglement remains…

Recent studies have revealed various manifestations of position bias in transformer architectures, from the "lost-in-the-middle" phenomenon to attention sinks, yet a comprehensive theoretical understanding of how attention masks and…

So far, expensive finetuning beyond the pretraining sequence length has been a requirement for effectively extending the context of language models (LM). In this work, we break this key bottleneck by Dropping the Positional Embeddings of…

Vision-and-language (V-L) tasks require the system to understand both vision content and natural language, thus learning fine-grained joint representations of vision and language (a.k.a. V-L representations) is of paramount importance.…

Recent advancements in Multimodal Large Language Models (MLLMs) have significantly enhanced cross-modal understanding and reasoning by incorporating Chain-of-Thought (CoT) reasoning in the semantic space. Building upon this, recent studies…

Multimodal Large Language Models frequently suffer from inference hallucinations, partially stemming from language priors dominating visual evidence. Existing training-free mitigation methods either perturb the visual representation and…

Length extrapolation algorithms based on Rotary position embedding (RoPE) have shown promising results in extending the context length of language models. However, understanding how position embedding can capture longer-range contextual…

Positional Encodings (PEs) are a critical component of Transformer-based Large Language Models (LLMs), providing the attention mechanism with important sequence-position information. One of the most popular types of encoding used today in…

Large Language Models (LLMs) are known to have limited extrapolation ability beyond their pre-trained context window, constraining their application in downstream tasks with lengthy inputs. Recent studies have sought to extend LLMs' context…

Training models with longer in-context lengths is a significant challenge for multimodal model due to substantial GPU memory and computational costs. This exploratory study does not present state-of-the-art models; rather, it introduces an…

Extending large language models (LLMs) to process longer inputs is crucial for a wide range of applications. However, the substantial computational cost of transformers and limited generalization of positional encoding restrict the size of…