English
Related papers

Related papers: The Impact of Positional Encoding on Length Genera…

200 papers

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…

Computation and Language · Computer Science 2025-05-14 Federico Barbero , Alex Vitvitskyi , Christos Perivolaropoulos , Razvan Pascanu , Petar Veličković

Positional encoding is essential for large language models (LLMs) to represent sequence order, yet recent studies show that Rotary Position Embedding (RoPE) can induce massive activation. We investigate the source of these instabilities via…

Computation and Language · Computer Science 2026-01-07 Jing Xiong , Liyang Fan , Hui Shen , Zunhai Su , Min Yang , Lingpeng Kong , Ngai Wong

Rotary Positional Embedding (RoPE) is a common choice in transformer architectures for encoding relative positional information. Although earlier work has examined omitting RoPE in specific layers, the effect of varying the fraction of…

Machine Learning · Computer Science 2026-03-13 Mohammad Aflah Khan , Krishna P. Gummadi , Manish Gupta , Abhilasha Ravichander

Currently, many studies view DNA sequences as a special type of language and utilize Transformers to model them. These studies use fixed-length k-mer segmentation and BPE subword tokenization but lack a systematic evaluation to determine…

Computation and Language · Computer Science 2025-07-22 Chenlei Gong , Yuanhe Tian , Lei Mao , Yan Song

Position encoding recently has shown effective in the transformer architecture. It enables valuable supervision for dependency modeling between elements at different positions of the sequence. In this paper, we first investigate various…

Computation and Language · Computer Science 2023-11-09 Jianlin Su , Yu Lu , Shengfeng Pan , Ahmed Murtadha , Bo Wen , Yunfeng Liu

Preventing the performance decay of Transformers on inputs longer than those used for training has been an important challenge in extending the context length of these models. Though the Transformer architecture has fundamentally no limits…

Transformer-based models have become the dominant paradigm for neural combinatorial optimization (NCO) of vehicle routing problems (VRPs), yet the role of positional encoding (PE) in these architectures remains largely unexplored. Unlike…

The distinguishing power of graph transformers is closely tied to the choice of positional encoding: features used to augment the base transformer with information about the graph. There are two primary types of positional encoding:…

Machine Learning · Computer Science 2024-08-26 Mitchell Black , Zhengchao Wan , Gal Mishne , Amir Nayyeri , Yusu Wang

Transformer architectures rely on position encodings to model the spatial structure of input data. Rotary Position Encoding (RoPE) is a widely used method in language models that encodes relative positions through fixed, block-diagonal,…

Computer Vision and Pattern Recognition · Computer Science 2025-08-19 Sophie Ostmeier , Brian Axelrod , Maya Varma , Michael E. Moseley , Akshay Chaudhari , Curtis Langlotz

We identify intrinsic limitations of Rotary Positional Embeddings (RoPE) in Transformer-based long-context language models. Our theoretical analysis abstracts away from the specific content of the context and depends only on its length. We…

Computation and Language · Computer Science 2026-05-18 Yufeng Du , Phillip Harris , Minyang Tian , Eliu A Huerta , Srikanth Ronanki , Subendhu Rongali , Aram Galstyan , Hao Peng

Many positional encodings (PEs) are designed to exhibit long-term decay, based on an entrenched and long-standing inductive opinion: tokens farther away from the current position carry less relevant information. We argue that long-term…

Computation and Language · Computer Science 2024-12-06 Yuhan Chen , Ang Lv , Jian Luan , Bin Wang , Wei Liu

We prove under practical assumptions that Rotary Positional Embedding (RoPE) introduces an intrinsic distance-dependent bias in attention scores that limits RoPE's ability to model long-context. RoPE extension methods may alleviate this…

Computation and Language · Computer Science 2026-05-12 Yu Wang , Sheng Shen , Rémi Munos , Hongyuan Zhan , Yuandong Tian

Length generalization, defined as the ability to extrapolate from shorter training sequences to longer test ones, is a significant challenge for language models. This issue persists even with large-scale Transformers handling relatively…

Machine Learning · Computer Science 2024-02-15 Yongchao Zhou , Uri Alon , Xinyun Chen , Xuezhi Wang , Rishabh Agarwal , Denny Zhou

Recent studies have demonstrated the effectiveness of position encoding in transformer architectures. By incorporating positional information, this approach provides essential guidance for modeling dependencies between elements across…

Machine Learning · Computer Science 2025-08-27 Avinash Amballa

Transformer architectures rely on explicit position encodings in order to preserve a notion of word order. In this paper, we argue that existing work does not fully utilize position information. For example, the initial proposal of a…

Computation and Language · Computer Science 2020-09-30 Zhiheng Huang , Davis Liang , Peng Xu , Bing Xiang

We propose a novel positional encoding for learning graph on Transformer architecture. Existing approaches either linearize a graph to encode absolute position in the sequence of nodes, or encode relative position with another node using…

Machine Learning · Computer Science 2022-10-17 Wonpyo Park , Woonggi Chang , Donggeon Lee , Juntae Kim , Seung-won Hwang

An important aspect subtending language understanding and production is the ability to independently encode positional and symbolic information of the words within a sentence. In Transformers, positional information is typically encoded…

Machine Learning · Computer Science 2025-11-18 Felipe Urrutia , Jorge Salas , Alexander Kozachinskiy , Cristian Buc Calderon , Hector Pasten , Cristobal Rojas

We conducted empirical experiments to assess the transferability of a light curve transformer to datasets with different cadences and magnitude distributions using various positional encodings (PEs). We proposed a new approach to…

A major challenge for transformers is generalizing to sequences longer than those observed during training. While previous works have empirically shown that transformers can either succeed or fail at length generalization depending on the…

Machine Learning · Computer Science 2025-05-01 Xinting Huang , Andy Yang , Satwik Bhattamishra , Yash Sarrof , Andreas Krebs , Hattie Zhou , Preetum Nakkiran , Michael Hahn

The Position Embedding (PE) is critical for Vision Transformers (VTs) due to the permutation-invariance of self-attention operation. By analyzing the input and output of each encoder layer in VTs using reparameterization and visualization,…

Computer Vision and Pattern Recognition · Computer Science 2022-12-23 Runyi Yu , Zhennan Wang , Yinhuai Wang , Kehan Li , Yian Zhao , Jian Zhang , Guoli Song , Jie Chen