metadata

dataset_info:
  features:
    - name: id
      dtype: int64
    - name: haystack
      dtype: string
    - name: keys
      dtype: string
    - name: values
      dtype: string
    - name: question
      dtype: string
    - name: context_length
      dtype: int64
    - name: num_kv_pairs
      dtype: int64
    - name: repeat_number
      dtype: int64
    - name: needle_depth
      dtype: string
    - name: num_needles
      dtype: int64
    - name: needle_placement
      dtype: string
    - name: conditional_character
      dtype: string
    - name: thread_length
      dtype: int64
    - name: thread_direction
      dtype: string
    - name: num_threads
      dtype: int64
  splits:
    - name: Single_Needle
      num_bytes: 159048560
      num_examples: 660
    - name: Multiple_Needles
      num_bytes: 261371340
      num_examples: 1080
    - name: Conditional_Needles
      num_bytes: 260974140
      num_examples: 1080
    - name: Single_Threads
      num_bytes: 564730140
      num_examples: 2340
    - name: Multi_Threads
      num_bytes: 1391847750
      num_examples: 5700
  download_size: 1798326219
  dataset_size: 2637971930
configs:
  - config_name: default
    data_files:
      - split: Single_Needle
        path: data/Single_Needle-*
      - split: Multiple_Needles
        path: data/Multiple_Needles-*
      - split: Conditional_Needles
        path: data/Conditional_Needles-*
      - split: Single_Threads
        path: data/Single_Threads-*
      - split: Multi_Threads
        path: data/Multi_Threads-*
license: mit
language:
  - en

Needle Threading: Can LLMs Follow Threads through Near-Million-Scale Haystacks?

Dataset Description

Homepage: https://needle-threading.github.io
Paper: Needle Threading: Can LLMs Follow Threads through Near-Million-Scale Haystacks?
Repository Needle Threading

Dataset Summary

As the context limits of Large Language Models (LLMs) increase, the range of possible applications and downstream functions broadens. Although the development of longer context models has seen rapid gains recently, our understanding of how effectively they use their context has not kept pace.

To address this, we conduct a set of retrieval experiments designed to evaluate the capabilities of 17 leading LLMs, such as their ability to follow threads of information through the context window.

Strikingly, we find that many models are remarkably thread-safe: capable of simultaneously following multiple threads without significant loss in performance. Still, for many models, we find the effective context limit is significantly shorter than the supported context length, with accuracy decreasing as the context window grows.

Example Usage

Option 1: HuggingFace datasets

from datasets import load_dataset

# task splits can be downloaded separately:
# splits = ['Single_Needle', 'Multi_Needle', 'Conditional_Needle', 'Single_Thread', 'Multi_Thread']
single_needle_dataset = load_dataset("jonathan-roberts1/needle-threading", split='Single_Needle')

"""
Dataset({
    features: ['id', 'haystack', 'keys', 'values', 'question', 'context_length', 'num_kv_pairs',
    'repeat_number', 'needle_depth', 'num_needles', 'needle_placement', 'conditional_character',
    'thread_length', 'thread_direction', 'num_threads'],
    num_rows: 660
})
Note the units of context_length are number of characters.
"""

# query individual questions
single_needle_dataset[5] # e.g., the 6th element
"""
{'id': 5, 'haystack': '{"e3e70682-c209-4cac-629f-6fbed82c07cd": "f728b4fa-4248-5e3a-0a5d-2f346baa9455",
"eb1...": "964a870c-7c87-9b74-1d87-8f9f9cdf5a86"}', 'keys': '247a8333-f7b0-b7d2-cda8-056c3d15eef7',
'values': '1759edc3-72ae-2244-8b01-63c1cd9d2b7d', 'question': 'Extract the value corresponding to
the specified key in the JSON object. Key: "247a83...-cda8-056c3d15eef7"\n Corresponding value: ',
'context_length': 2000, 'num_kv_pairs': 25, 'repeat_number': 0, 'needle_depth': '50', 'num_needles': 1,
'needle_placement': 'depthwise', 'conditional_character': 'N/A', 'thread_length': 1,
'thread_direction': 'N/A', 'num_threads': 0}
"""

Option 2: Manual download

Directly downloading image files and question data from the needle-threading HuggingFace repository into the data directory in this repo.

cd data
wget https://huggingface.co/datasets/jonathan-roberts1/needle-threading/resolve/main/data_json.zip
unzip data_json.zip && rm data_json.zip

Expected structure

├── data
    ├── json_data
        ├── Single_Needle.json
        ├── Multiple_Needles.json
        ├── Conditional_Needles.json
        ├── Single_Threads.json
        ├── Multi_Threads.json

Note: data_json/ needs to be downloaded.

Please visit our GitHub repository for example inference code.

Dataset Curators

This dataset was curated by Jonathan Roberts, Kai Han, and Samuel Albanie

Citation

If you found our work useful in your own research, please consider citing our paper:

@article{roberts2024needle,
  title={Needle Threading: Can LLMs Follow Threads through Near-Million-Scale Haystacks?},
  author={Roberts, Jonathan and Han, Kai and Albanie, Samuel},
  journal={arXiv preprint arXiv:2411.05000},
  year={2024}
}