RULER: What's the Real Context Size of Your Long-Context Language Models?
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_ | 1866916349064773632 |
|---|---|
| author | Hsieh, Cheng-Ping Sun, Simeng Kriman, Samuel Acharya, Shantanu Rekesh, Dima Jia, Fei Zhang, Yang Ginsburg, Boris |
| author_facet | Hsieh, Cheng-Ping Sun, Simeng Kriman, Samuel Acharya, Shantanu Rekesh, Dima Jia, Fei Zhang, Yang Ginsburg, Boris |
| contents | The needle-in-a-haystack (NIAH) test, which examines the ability to retrieve a piece of information (the "needle") from long distractor texts (the "haystack"), has been widely adopted to evaluate long-context language models (LMs). However, this simple retrieval-based test is indicative of only a superficial form of long-context understanding. To provide a more comprehensive evaluation of long-context LMs, we create a new synthetic benchmark RULER with flexible configurations for customized sequence length and task complexity. RULER expands upon the vanilla NIAH test to encompass variations with diverse types and quantities of needles. Moreover, RULER introduces new task categories multi-hop tracing and aggregation to test behaviors beyond searching from context. We evaluate 17 long-context LMs with 13 representative tasks in RULER. Despite achieving nearly perfect accuracy in the vanilla NIAH test, almost all models exhibit large performance drops as the context length increases. While these models all claim context sizes of 32K tokens or greater, only half of them can maintain satisfactory performance at the length of 32K. Our analysis of Yi-34B, which supports context length of 200K, reveals large room for improvement as we increase input length and task complexity. We open source RULER to spur comprehensive evaluation of long-context LMs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_06654 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | RULER: What's the Real Context Size of Your Long-Context Language Models? Hsieh, Cheng-Ping Sun, Simeng Kriman, Samuel Acharya, Shantanu Rekesh, Dima Jia, Fei Zhang, Yang Ginsburg, Boris Computation and Language The needle-in-a-haystack (NIAH) test, which examines the ability to retrieve a piece of information (the "needle") from long distractor texts (the "haystack"), has been widely adopted to evaluate long-context language models (LMs). However, this simple retrieval-based test is indicative of only a superficial form of long-context understanding. To provide a more comprehensive evaluation of long-context LMs, we create a new synthetic benchmark RULER with flexible configurations for customized sequence length and task complexity. RULER expands upon the vanilla NIAH test to encompass variations with diverse types and quantities of needles. Moreover, RULER introduces new task categories multi-hop tracing and aggregation to test behaviors beyond searching from context. We evaluate 17 long-context LMs with 13 representative tasks in RULER. Despite achieving nearly perfect accuracy in the vanilla NIAH test, almost all models exhibit large performance drops as the context length increases. While these models all claim context sizes of 32K tokens or greater, only half of them can maintain satisfactory performance at the length of 32K. Our analysis of Yi-34B, which supports context length of 200K, reveals large room for improvement as we increase input length and task complexity. We open source RULER to spur comprehensive evaluation of long-context LMs. |
| title | RULER: What's the Real Context Size of Your Long-Context Language Models? |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2404.06654 |