Self-Consistency Falls Short! The Adverse Effects of Positional Bias on Long-Context Problems
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909819343994880 |
|---|---|
| author | Byerly, Adam Khashabi, Daniel |
| author_facet | Byerly, Adam Khashabi, Daniel |
| contents | Self-consistency (SC) improves the performance of large language models (LLMs) across various tasks and domains that involve short content. However, does this support its effectiveness for long-context problems?
We challenge the assumption that SC's benefits generalize to long-context settings, where LLMs often struggle with position bias, the systematic over-reliance on specific context regions-which hinders their ability to utilize information effectively from all parts of their context. Through comprehensive experimentation with varying state-of-the-art models, tasks, and SC formulations, we find that SC not only fails to improve but actively degrades performance on long-context tasks. This degradation is driven by persistent position bias, which worsens with longer context lengths and smaller model sizes but remains invariant to prompt format or task type. Unlike short-context tasks, where SC diversifies reasoning paths, long-context SC amplifies positional errors. These comprehensive results provide valuable insight into the limitations of current LLMs in long-context understanding and highlight the need for more sophisticated approaches. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2411_01101 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Self-Consistency Falls Short! The Adverse Effects of Positional Bias on Long-Context Problems Byerly, Adam Khashabi, Daniel Computation and Language Self-consistency (SC) improves the performance of large language models (LLMs) across various tasks and domains that involve short content. However, does this support its effectiveness for long-context problems? We challenge the assumption that SC's benefits generalize to long-context settings, where LLMs often struggle with position bias, the systematic over-reliance on specific context regions-which hinders their ability to utilize information effectively from all parts of their context. Through comprehensive experimentation with varying state-of-the-art models, tasks, and SC formulations, we find that SC not only fails to improve but actively degrades performance on long-context tasks. This degradation is driven by persistent position bias, which worsens with longer context lengths and smaller model sizes but remains invariant to prompt format or task type. Unlike short-context tasks, where SC diversifies reasoning paths, long-context SC amplifies positional errors. These comprehensive results provide valuable insight into the limitations of current LLMs in long-context understanding and highlight the need for more sophisticated approaches. |
| title | Self-Consistency Falls Short! The Adverse Effects of Positional Bias on Long-Context Problems |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2411.01101 |