Get the app
extract 0 just added trending

openrae/duels

published by @openrae

Alignment dataset tracked in the /datasets sub-catalog.

Lifetime downloads
83
Last 30 days
83
Likes
5
Size
356.5 MB
Created on HF
2026-10-18
Age
3 days ago

Description

OpenRÆ Duels: BSidesNYC 2026 Red–Blue Agent Traces

Paired red-team and blue-team coding-agent traces collected during live, supervised cyber-range duels at the BSidesNYC 2026 AI Security Village in the public OpenRÆ TechVault environment.

This release contains 40 paired duels and 80 role traces. It preserves the collector recordings byte for byte under source/episodes/, provides trace-viewer derivatives under viewer/, and exposes linked Parquet tables under data/.… See the full description on the dataset page: https://huggingface.co/datasets/openrae/duels.

Tags

task_categories:text-generationlanguage:enlicense:cc-by-4.0size_categories:100K<n<1Mformat:parquetmodality:tabularmodality:textlibrary:datasetslibrary:pandaslibrary:polarslibrary:mlcroissantregion:usagent-tracesagentscybersecurityred-teamblue-teamoffensive-aidefensive-aitool-useopenrae

README current version from Hugging Face

pretty_name: 'OpenRÆ Duels: BSidesNYC 2026 Red–Blue Agent Traces'
license: cc-by-4.0
language:

  • en
    task_categories:
  • text-generation
    tags:
  • agent-traces
  • agents
  • cybersecurity
  • red-team
  • blue-team
  • offensive-ai
  • defensive-ai
  • tool-use
  • openrae
    configs:
  • config_name: episodes
    default: true
    data_files:
    • split: full
      path: data/episodes.parquet
  • config_name: traces
    data_files:
    • split: full
      path: data/traces.parquet
  • config_name: messages
    data_files:
    • split: full
      path: data/messages.parquet
  • config_name: events
    data_files:
    • split: full
      path: data/events.parquet
  • config_name: director_records
    data_files:
    • split: full
      path: data/director_records.parquet
  • config_name: artifacts
    data_files:
    • split: full
      path: data/artifacts.parquet

OpenRÆ Duels: BSidesNYC 2026 Red–Blue Agent Traces

Paired red-team and blue-team coding-agent traces collected during live, supervised cyber-range duels at the BSidesNYC 2026 AI Security Village in the public OpenRÆ TechVault environment.

This release contains 40 paired duels and 80 role traces. It preserves the collector recordings byte for byte under source/episodes/, provides trace-viewer derivatives under viewer/, and exposes linked Parquet tables under data/.

Structure

  • episodes: one row per paired duel.
  • traces: one row per Red or Blue trace.
  • messages: semantic reasoning, text, tool calls, and tool results.
  • events: every recorded collector-envelope event, including lifecycle and token telemetry.
  • director_records: contemporaneous experimental notes and interventions.
  • artifacts: hashes and roles for package artifacts.

Use the full split. Curation status is a column because this observational corpus includes qualified, partial, research-only, and unassessed sessions.

Source and provenance

The agents ran inside isolated OpenRÆ cyber ranges using the public lilrae platform and public TechVault scenario. See docs/PROVENANCE.md, docs/COLLECTION.md, and docs/SCHEMA.md.

Citation and license

README history 5 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. 2026-10-18Add agents and AI role tags65789722.2 KB
    Loading...
  2. 2026-10-18Upload OpenRÆ Duels v1.0.0 dataset (part 3)9b0cc1a2.2 KB
    Loading...
  3. 2026-10-18Upload OpenRÆ Duels v1.0.0 dataset (part 2)ae45f132.2 KB
    Loading...
  4. 2026-10-18Upload OpenRÆ Duels v1.0.0 dataset30440bc2.2 KB
    Loading...
  5. 2026-10-18initial commit0deb4800 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.

Related datasets same stage · extract

Anthropic/hh-rlhf
2.1M dl 2442 likes 90.4 MB
Dataset Card for HH-RLHF Dataset Summary This repository provides access to two different kinds of data: Human ...
djain95/sae-jailbreaks-cache
468.6K dl 1 likes 30.6 GB
mosaicml/dolly_hhrlhf
309.5K dl 113 likes 23.7 MB
Dataset Card for "dolly_hhrlhf" This dataset is a combination of Databrick's dolly-15k dataset and a filtered subset of Anthrop...
Dahoas/full-hh-rlhf
266.3K dl 104 likes 129.8 MB
Dataset Card for "full-hh-rlhf" Anthropic's HH dataset reformatted into prompt, chosen, rejected samples.
mlabonne/harmful_behaviors
198.2K dl 169 likes 23 KB
mlabonne/harmless_alpaca
190.2K dl 51 likes 1.2 MB
walledai/AdvBench
176.3K dl 123 likes 38 KB
Dataset Card for AdvBench Paper: Universal and Transferable Adversarial Attacks on Aligned Language Models Data: AdvBench Datas...
JailbreakV-28K/JailBreakV-28k
116.4K dl 80 likes 288.8 MB
⛓‍💥 JailBreakV-28K: A Benchmark for Assessing the Robustness of MultiModal Large Language Models against Jailbreak Attacks 🌐 ...

Read further

Community best practices (extraction)
5 rules the field converged on for building the pair.
Building your own extraction pair
When the reference pair is not enough. Sizing, framing, and evaluation.
Datasets that shaped abliteration
The five corpora that did most of the work.
Non-English abliteration communities
How Chinese, Russian, Japanese, and Korean practitioners build extraction pairs.

Metadata

License
cc-by-4.0

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