Skip to content

BeaverTails is a collection of datasets designed to facilitate research on safety alignment in large language models (LLMs).

active 2023-08-162026-04-14 (UTC)

Partial coverage26,158 / 26,322 hourly files (99%) · 2 absent upstream · 163 failed, retryable2023-08-152026-08-15 (UTC)— sampled evenly across the window, so rankings and trends hold; absolute counts scale up.
Events
161
Pushes
1
Pull requests
2
Issues
10
Stars
137
Forks
4

Activity over time

Daily event counts in the loaded window

Line chart, 973 days from 2023-08-16 to 2026-04-14. Pushes: 1 total, peak 1 in a day. Pull requests: 2 total, peak 1 in a day. Issues: 10 total, peak 1 in a day. Comments: 5 total, peak 3 in a day. Stars: 137 total, peak 3 in a day.

  • Pushes
  • Pull requests
  • Issues
  • Comments
  • Stars

Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.

Top contributors

Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity

ContributorContributionsPushesPRsComments
mickelliu9124
hxhcreate2000
zmsn-20772001
HuengchI1000
calico-12261000
nonstopfor1000
LxzGordon1000
yananchen19891000
ChanLiang1000

Recent activity

Latest issues, pull requests and releases

  • Issue#10ChanLiang2025-01-14 07:21
    [Question] Seeking the Helpfulness evaluation model
  • Issue#9yananchen19892024-07-23 18:01
    [Question] differences among two datasets
  • Pull request#5mickelliu2024-06-20 00:29
  • Issue#7HuengchI2024-05-17 11:59
    [Question] 关于QAModeration的num_labels的问题
  • Issue#6hxhcreate2024-02-26 05:14
    [Question] I was wondering the difference between Beavertails dataset and SafeRLHF dataset in huggingface on the datasource aspect
  • Issue#6hxhcreate2024-01-16 14:39
    [Question] I was wondering the difference between Beavertails dataset and SafeRLHF dataset in huggingface on the datasource aspect
  • Issue comment#3mickelliu2023-10-31 06:23
    [Question] 关于总数据量的问题
  • Issue#3mickelliu2023-10-31 06:23
    [Question] 关于总数据量的问题
  • Pull request#5mickelliu2023-10-27 19:47
  • Issue comment#4mickelliu2023-10-27 19:36
    [BUG] Duplicate responses lead to training collapse
  • Issue#4mickelliu2023-10-27 19:36
    [BUG] Duplicate responses lead to training collapse
  • Issue comment#4mickelliu2023-10-27 18:39
    [BUG] Duplicate responses lead to training collapse
  • Issue comment#3mickelliu2023-10-27 18:30
    [Question] 关于总数据量的问题
  • Issue#4LxzGordon2023-10-25 04:12
    [BUG] 一个小bug
  • Issue#3nonstopfor2023-10-19 01:46
    [Question] 关于总数据量的问题
  • Issue#2zmsn-20772023-09-16 14:12
    [Question] QA Moderation是否支持中文模型
  • Issue comment#2zmsn-20772023-09-14 17:55
    [Question]

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 137 stars here means stars gained during the window, not the repo's star count.