Get the app
misc 0 0 0

agentlans/chat-annotated

published by @agentlans

Alignment dataset tracked in the /datasets sub-catalog.

Lifetime downloads
2K
Last 30 days
153
Likes
3
Size
208.3 MB
Created on HF
2025-12-28
Age
9 months ago

Description

Annotated Human-AI Chats

Variable

Mean

Standard Deviation

safety

6.5310

1.0179

quality

7.8994

1.9667

difficulty

6.6519

2.1795

There are two columns for refusal:

refusal: the categories from agentlans/chat-refusal

refusal_classifier: the output of

agentlans/snowflake-arctic-embed-xs-refusal-classifier for English conversations

agentlans/multilingual-e5-small-refusal-classifier for non-English conversations

The accept column is true if safety > 5, quality >… See the full description on the dataset page: https://huggingface.co/datasets/agentlans/chat-annotated.

Tags

task_categories:text-classificationtask_categories:text-rankinglanguage:enlanguage:zhlanguage:rulanguage:arlanguage:eslanguage:frlanguage:vilanguage:idlanguage:ptlanguage:jalanguage:falanguage:delanguage:kolanguage:trlanguage:itlanguage:pllanguage:talanguage:silanguage:uklanguage:rolicense:cc-by-4.0size_categories:100K<n<1Mformat:jsonmodality:tabularmodality:textlibrary:datasetslibrary:pandaslibrary:polarslibrary:mlcroissantregion:us

README current version from Hugging Face

license: cc-by-4.0
configs:

  • config_name: main
    data_files:
    • path:
      • annot.jsonl.zst
        split: train
        default: true
  • config_name: difficulty
    data_files:
    • path:
      • difficulty_train.jsonl.zst
        split: train
    • path:
      • difficulty_test.jsonl.zst
        split: test
  • config_name: safety
    data_files:
    • path:
      • safety_train.jsonl.zst
        split: train
    • path:
      • safety_test.jsonl.zst
        split: test
  • config_name: quality
    data_files:
    • path:
      • quality_train.jsonl.zst
        split: train
    • path:
      • quality_test.jsonl.zst
        split: test
  • config_name: refusal
    data_files:
    • path:
      • refusal_train.jsonl.zst
        split: train
    • path:
      • refusal_test.jsonl.zst
        split: test
  • config_name: accept
    data_files:
    • path:
      • accept_train.jsonl.zst
        split: train
    • path:
      • accept_test.jsonl.zst
        split: test
        language:
  • en
  • zh
  • ru
  • ar
  • es
  • fr
  • vi
  • id
  • pt
  • ja
  • fa
  • de
  • ko
  • tr
  • it
  • pl
  • ta
  • si
  • uk
  • ro
    task_categories:
  • text-classification
  • text-ranking

Annotated Human-AI Chats

Variable Mean Standard Deviation
safety 6.5310 1.0179
quality 7.8994 1.9667
difficulty 6.6519 2.1795

There are two columns for refusal:

The accept column is true if safety > 5, quality > 6, and refusal_classifier is "Non-refusal".

Language Number of rows
en 50000
zh 12372
ru 11229
ar 3207
es 3138
fr 2898
vi 2674
id 2570
pt 2192
ja 1237
fa 1008
de 957
ko 802
tr 770
it 610
pl 392
ta 297
si 266
uk 223
ro 206
Source Count
allenai/WildChat 36508
QuixiAI/dolphin-distill 16363
arcee-ai/The-Tome 8688
allenai/tulu-3-sft-mixture 8375
Locutusque/hercules-v6.9 7501
BAAI/Infinity-Instruct 7M_core 7411
mlabonne/open-perfectblend 6789
NousResearch/Hermes-3-Dataset 5949
agentlans/DSULT-Core-ShareGPT-X 1473
openchat/openchat_sharegpt4_dataset 938

README history 18 revisions

Every night we snapshot the README of every dataset in the catalog. When the SHA changes we archive the new version and diff it against the last. This is the evolving thought record of the alignment-data field: what the author decided to say about the corpus, and how that framing shifted over time.

  1. 2026-01-17Update README.mde3505893.7 KB
    Loading...
  2. 2026-01-17Upload 2 files9916c723.5 KB
    Loading...
  3. 2026-01-17Upload annot.jsonl.zst34275c03.5 KB
    Loading...
  4. 2026-01-17Upload annot.jsonl.zstfc74fc23.5 KB
    Loading...
  5. 2026-01-17Update README.mdd36b6743.5 KB
    Loading...
  6. 2026-01-17Update README.md87c2edd3.3 KB
    Loading...
  7. 2026-01-17Upload annot.jsonl.zstf3764c23.2 KB
    Loading...
  8. 2026-01-17Update README.mde2de1e83.2 KB
    Loading...
  9. 2026-01-17Upload annot.jsonl.zst4ddc1da2.9 KB
    Loading...
  10. 2025-12-31Upload 3 filesddda86f2.9 KB
    Loading...
  11. 2025-12-31Update README.md4ce3c912.9 KB
    Loading...
  12. 2025-12-31Upload 3 filesb1da1a72.8 KB
    Loading...
  13. 2025-12-28Update README.md09b26592.8 KB
    Loading...
  14. 2025-12-28Upload 7 filesf5cdcee2.8 KB
    Loading...
  15. 2025-12-28Upload 7 filesfcb3e092.8 KB
    Loading...
  16. 2025-12-28Upload 7 files243c1812.8 KB
    Loading...
  17. 2025-12-28Update README.md24026392.8 KB
    Loading...
  18. 2025-12-28initial commitb33872e30 B
    Loading...

Used by abliterated models none yet

No abliterated models in our catalog have declared this dataset in their YAML metadata yet. This can mean the dataset is used but not declared, is used outside abliteration workflows, or is new. The link will populate automatically as models are indexed.

Community discussions 1 thread · 1 open · 1 comments

Threads posted by the community on Hugging Face: bug reports, questions about usage, pull requests on the README, license clarifications. Click any thread to expand and read the full conversation inline. Newest first.

  1. 2025-12-29[bot] Conversion to Parquetopen1 💬#1
    Loading...

Continue on Hugging Face →

Related datasets same stage · misc

nvidia/Aegis-AI-Content-Safety-Dataset-2.0
99.3K dl 114 likes 27.1 MB
🛡️ Nemotron Content Safety Dataset V2 The Nemotron Content Safety Dataset V2, formerly known as Aegis AI Content Safety Data...
allenai/tulu-3-sft-personas-instruction-following
88.4K dl 70 likes 37.4 MB
Dataset Descriptions This dataset contains 29980 examples and is synthetically created to enhance model's capabilities to follo...
meta-llama/Llama-3.1-8B-Instruct-evals
73.2K dl 36 likes 232.9 MB
Dataset Card for Llama-3.1-8B-Instruct Evaluation Result Details This dataset contains the Meta evaluation result details for...
Manusagents/GPT-5.5-Gemini-3.1-Pro-Grok-4-Claude-Fable-5-Mythos-5-Qwen-3.7-Max-and-more-Distillation-Dataset
58.3K dl 226 likes 10.8 GB
📖 The Open Distillation Codex 🌌 The Ultimate Open-Source Distillation Dataset — No Skip, Full, with Attack & Defe...
nvidia/Aegis-AI-Content-Safety-Dataset-1.0
36.1K dl 63 likes 5.2 MB
🛡️ Nemotron Content Safety Dataset V1 Nemotron Content Safety Dataset V1, formerly known as Aegis AI Content Safety Dataset, i...
innodatalabs/rt-gsm8k-gaia
35.9K dl 15 KB
rt-gsm8k dataset Red teaming gsm8k-gaia dataset. Generated from the "GSM8K" dataset and "GAIA" dataset. Dataset Str...
HuggingFaceH4/hhh_alignment
35.3K dl 27 likes 266 KB
This task evaluates language models on alignment, broken down into categories of helpfulness, honesty/accuracy, harmlessness, and ...
ethz-spylab/rlhf_trojan_dataset
26.2K dl 6 likes 29.9 MB
Dataset for the competition This is the official dataset for the competition "Find the Trojan: Universal Backdoor Detection in ...

Read further

Community best practices (architecture)
4 rules for MoE, hybrid SSM, FP8 requantization, and Gemma quirks.
The six workflow stages
What misc is doing more work than its name suggests.
Beyond English: multilingual corpora
The four language communities and their workarounds.
Non-English abliteration communities
Deep dive: huihui-ai, layer-window, slop-framing, CJK tokenizer fixes.

Metadata

License
cc-by-4.0

Repository

FILE .gitattributes FILE accept_test.jsonl.zst FILE accept_train.jsonl.zst FILE annot.jsonl.zst FILE difficulty_test.jsonl.zst FILE difficulty_train.jsonl.zst FILE quality_test.jsonl.zst FILE quality_train.jsonl.zst FILE README.md FILE refusal_test.jsonl.zst FILE refusal_train.jsonl.zst FILE safety_test.jsonl.zst FILE safety_train.jsonl.zst

File contents live on Hugging Face. Click a file or folder to open it there.

Download the dataset

0

Copy any of these snippets into your notebook or terminal. All three fetch directly from Hugging Face using your own credentials.

Recommended: datasets library (Python) 
from datasets import load_dataset
ds = load_dataset("agentlans/chat-annotated")
Raw snapshot (Python) 
from huggingface_hub import snapshot_download
snapshot_download(repo_id="agentlans/chat-annotated", repo_type="dataset")
git clone (requires git-lfs) 
git clone https://huggingface.co/datasets/agentlans/chat-annotated
Browse repository on Hugging Face ->
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Get the app Browse the catalog
Open in Abliteration