README.md

metadata

license: cc-by-nc-nd-4.0
tags:
  - cybersecurity
  - cti
  - yara
  - contrastive-learning
  - sentence-similarity
  - dual-encoder
library_name: transformers
base_model: sentence-transformers/all-mpnet-base-v2

FALCON dual-encoder — YARA / all-mpnet-base-v2

Contrastive encoder fine-tuned to map CTI text and YARA rules into a shared embedding space. Backbone: sentence-transformers/all-mpnet-base-v2.

Test-set metrics

split	recall@1	F1	threshold	diag mean	off-diag mean
pretrained	0.7325	0.2365	0.6850	0.7315	0.4856
run_0	0.9498	0.9214	0.6746	0.8456	0.0140
run_1	0.9509	0.9392	0.6951	0.9041	0.0207
run_2	0.9509	0.9351	0.6849	0.8838	0.0123
run_3	0.9509	0.9336	0.6979	0.9185	0.0234
run_4	0.9509	0.9412	0.7060	0.9625	0.0046

Training

Symmetric InfoNCE / NT-Xent over in-batch negatives. Best checkpoint selected by validation loss.

• Run 0 — batch=16, epochs=5, lr=2e-05, schedule=constant, T=0.05
• Run 1 — batch=50, epochs=10, lr=2e-05, schedule=constant, T=0.05
• Run 2 — batch=70, epochs=30, lr=2e-05, schedule=constant, T=0.05
• Run 3 — batch=128, epochs=30, lr=5e-05, schedule=warmup_cosine, T=0.05
• Run 4 — batch=70, epochs=50, lr=2e-05, schedule=constant, T=0.07

Loading

from transformers import AutoModel, AutoTokenizer
tok   = AutoTokenizer.from_pretrained("shaswatamitra/falcon-yara-dual-all-mpnet-base-v2", subfolder='rule')
model = AutoModel.from_pretrained("shaswatamitra/falcon-yara-dual-all-mpnet-base-v2", subfolder='rule')

Dual-encoder layout: this repo has rule/ (encodes YARA rules) and cti/ (encodes CTI text) subfolders. Load each with subfolder=....

Citation

@article{mitra2025falcon,
  title={FALCON: Autonomous Cyber Threat Intelligence Mining with LLMs for IDS Rule Generation},
  author={Mitra, Shaswata and Bazarov, Azim and Duclos, Martin and Mittal, Sudip and Piplai, Aritran and Rahman, Md Rayhanur and Zieglar, Edward and Rahimi, Shahram},
  journal={arXiv preprint arXiv:2508.18684},
  year={2025}
}