extract 0 0 0

Graph-COM/MTID

published by @Graph-COM

Alignment dataset tracked in the /datasets sub-catalog.

Lifetime downloads
1.2K
Last 30 days
202
Likes
2
Size
94.0 MB
Created on HF
2026-05-13
Age
5 months ago

Description

TurnGate: Response-Aware Defense Against Hidden Malicious Intent in Multi-Turn Dialogue

Overview

TurnGate is a response-aware defense mechanism designed to detect and mitigate hidden malicious intent in multi-turn dialogue systems. Defending state-of-the-art multi-turn malicious attacks like CKA-Agent.

MTID Dataset

We include the MTID (Multi-Turn Intent Dataset) here. This dataset contains a collection of… See the full description on the dataset page: https://huggingface.co/datasets/Graph-COM/MTID.

Tags

task_categories:text-classificationlanguage:enlicense:agpl-3.0size_categories:10K<n<100Kformat:jsonmodality:tabularmodality:textlibrary:datasetslibrary:pandaslibrary:polarslibrary:mlcroissantarxiv:2605.05630region:usSafetyDefenseJailbreakMulti-turnHarmfulBenign

README current version from Hugging Face


language:

  • en
    license: agpl-3.0
    size_categories:
  • 10K<n<100K
    task_categories:
  • text-classification
    pretty_name: MTID
    tags:
  • Safety
  • Defense
  • Jailbreak
  • Multi-turn
  • Harmful
  • Benign
    configs:
  • config_name: benign
    data_files:
    • split: train
      path: benign_train.jsonl
    • split: validation
      path: benign_valid.jsonl
    • split: test
      path: benign_test.jsonl
  • config_name: harmful
    data_files:
    • split: train
      path: harmful_train.jsonl
    • split: validation
      path: harmful_valid.jsonl
    • split: test
      path: harmful_test.jsonl

TurnGate: Response-Aware Defense Against Hidden Malicious Intent in Multi-Turn Dialogue

arXiv Website GitHub code Cite Python

Overview

TurnGate is a response-aware defense mechanism designed to detect and mitigate hidden malicious intent in multi-turn dialogue systems. Defending state-of-the-art multi-turn malicious attacks like CKA-Agent.

TurnGate Pipeline

MTID Dataset

We include the MTID (Multi-Turn Intent Dataset) here. This dataset contains a collection of multi-turn interactions focused on evaluating and training defenses against correlated knowledge attacks.

Dataset Structure

The dataset is split into train, valid, and test sets for both benign and harmful categories:

  • Total Unique Samples: 800 (400 Benign, 400 Harmful)
  • Rollouts per Sample: 20 (Total of 16,000 trajectories)
  • Format: Each line is a JSON object representing a single rollout.

The Hugging Face dataset viewer exposes two configs, benign and harmful, because the two categories contain category-specific annotation fields.

Cite

If you find this repository useful for your research, please consider citing the following paper:

@misc{shen2026turnlateresponseawaredefense,
      title={One Turn Too Late: Response-Aware Defense Against Hidden Malicious Intent in Multi-Turn Dialogue}, 
      author={Xinjie Shen and Rongzhe Wei and Peizhi Niu and Haoyu Wang and Ruihan Wu and Eli Chien and Bo Li and Pin-Yu Chen and Pan Li},
      year={2026},
      eprint={2605.05630},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2605.05630}, 
}

README history 7 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. 2026-09-07fix meta_intent (#3)9e9f7243 KB
    Loading...
  2. 2026-05-17Fix Hugging Face dataset configs for mixed schemas9dc74b43 KB
    Loading...
  3. 2026-05-13Add task category to metadata (#1)54ab7ca2.5 KB
    Loading...
  4. 2026-05-13fix image4687edc2.5 KB
    Loading...
  5. 2026-05-13Update dataset meta95fdd902.4 KB
    Loading...
  6. 2026-05-13init commit1a68fba2.3 KB
    Loading...
  7. 2026-05-13initial commit9c8759a26 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 3 threads · 1 open · 4 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. 2026-09-07PRfix meta_intentmerged1 💬#3
    Loading...
  2. 2026-05-18[bot] Conversion to Parquetopen1 💬#2
    Loading...
  3. 2026-05-13PRAdd task category to metadatamerged2 💬#1
    Loading...

Related datasets same stage · extract

Read further

Metadata

License
agpl-3.0

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