evaluate 0 0 0

PhilippBlue/mcq_safety_boxed

published by @PhilippBlue

Alignment dataset tracked in the /datasets sub-catalog.

Lifetime downloads
65
Last 30 days
9
Likes
-
Size
12.1 MB
Created on HF
2026-06-10
Age
4 months ago

Description

mcq_safety_boxed

Safety post-training data for Qwen3-1.7B

(Qwen/Qwen3-1.7B), CS-552 (EPFL,

Spring 2026), team ma-que (group 11). Released for reproducibility.

Files

File

Examples

train.jsonl

15,175

Reformatted boxed-letter MCQ

Real-world safety multiple-choice questions reformatted to a single-answer

\boxed{LETTER} schema. Derived from

cs-552-2026-vibe-trainers/mcq_safety (built from

SafetyBench, SALAD-Bench, and WildGuardMix)… See the full description on the dataset page: https://huggingface.co/datasets/PhilippBlue/mcq_safety_boxed.

Tags

task_categories:multiple-choicelanguage:enlicense:cc-by-4.0size_categories:10K<n<100Kformat:jsonmodality:textlibrary:datasetslibrary:pandaslibrary:polarslibrary:mlcroissantregion:ussafetymcqqwen3

README current version from Hugging Face


license: cc-by-4.0
task_categories:

  • multiple-choice
    language:
  • en
    tags:
  • safety
  • mcq
  • qwen3
    pretty_name: ma-que Safety Data

mcq_safety_boxed

Safety post-training data for Qwen3-1.7B
(Qwen/Qwen3-1.7B), CS-552 (EPFL,
Spring 2026), team ma-que (group 11). Released for reproducibility.

Files

File Examples
train.jsonl 15,175

Reformatted boxed-letter MCQ

Real-world safety multiple-choice questions reformatted to a single-answer
\boxed{LETTER} schema. Derived from
cs-552-2026-vibe-trainers/mcq_safety (built from
SafetyBench, SALAD-Bench, and WildGuardMix), filtered to standard A–D answers.
Each line is a chat record:

{"messages": [
    {"role": "system",    "content": "You are a safety expert ... \\boxed{A}. /no_think"},
    {"role": "user",      "content": "<question with A) ... B) ... options>"},
    {"role": "assistant", "content": "\\boxed{B}"}],
 "answer": "B", "source": "safetybench"}

Provenance & licensing

Upstream sources retain their original licences (SafetyBench, SALAD-Bench,
WildGuardMix, PKU-SafeRLHF). This repository redistributes reformatted
derivatives and/or model-generated reasoning traces; verify upstream licences
before downstream use. Model-generated traces may inherit factual errors and the
Western-centric framing of the teacher model and source benchmarks and are not
authoritative safety guidance. Reproduce with push_safety_data.py.

README history 2 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. 2026-06-10Upload folder using huggingface_hubfbca2d71.6 KB
    Loading...
  2. 2026-06-10initial commit151d1090 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. 2026-06-11[bot] Conversion to Parquetopen1 💬#1
    Loading...

Related datasets same stage · evaluate

Read further

Metadata

License
cc-by-4.0

Repository

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("PhilippBlue/mcq_safety_boxed")
Raw snapshot (Python)
from huggingface_hub import snapshot_download
snapshot_download(repo_id="PhilippBlue/mcq_safety_boxed", repo_type="dataset")
git clone (requires git-lfs)
git clone https://huggingface.co/datasets/PhilippBlue/mcq_safety_boxed
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.

Open in Abliteration