metadata
dataset_info:
features:
- name: corpus_id
dtype: string
- name: arxiv_id
dtype: string
- name: date
dtype: string
- name: title
dtype: string
- name: abstract
dtype: string
- name: categories
list: string
- name: roles
list: string
- name: key_references
list:
- name: corpus_id
dtype: string
- name: num_citations
dtype: float64
- name: authors
list:
- name: author_id
dtype: string
- name: h_index
dtype: float64
- name: name
dtype: string
- name: num_citations
dtype: float64
- name: num_papers
dtype: float64
- name: publication_history
list: string
- name: citation_trajectory
list: int64
- name: specter2_embed
list: float64
- name: primary_cluster
dtype: float64
- name: soft_membership
list: float64
- name: novelty
dtype: float64
- name: citation_percentile
dtype: float64
splits:
- name: train
num_bytes: 1163322702
num_examples: 373716
- name: test
num_bytes: 1429513331
num_examples: 464942
download_size: 1815001585
dataset_size: 2592836033
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*