Cybersecurity cross-encoder
According to the model card, the model computes pairwise similarity scores between two texts for reranking, semantic search, and cybersecurity intelligence retrieval.
Open Source Model Profile · cisco-ai
SecureBERT2.0-cross_encoder
SecureBERT2.0-cross-encoder is a 149.6M-parameter ModernBERT cross-encoder from cisco-ai. According to the model card, it is a cybersecurity-domain fine-tune of SecureBERT 2.0 that scores pairwise text similarity for reranking and retrieval.
- Publisher
- cisco-ai
- Task
- sentence-similarity
- Model type
- modernbert
- License
- apache-2.0
- Library
- sentence-transformers
- Publication status
- Accepted · not indexed
Model overview
SecureBERT2.0-cross-encoder is published by cisco-ai as a sentence-similarity cross-encoder. The captured configuration identifies ModernBertForSequenceClassification with a modernbert model type, and Safetensors metadata reports 149605633 parameters. According to the model card, it was fine-tuned from cisco-ai/SecureBERT2.0-base for cybersecurity text similarity, reranking, and intelligence retrieval.
Recorded capabilities
ModernBERT sequence-classification base
Captured config identifies ModernBertForSequenceClassification with 149605633 parameters and sentence-transformers library support.
1024-token sequence length
According to the model card, maximum sequence length is 1024 tokens with a single similarity-score output label for English input.
Documented scope limits
The model card places generic non-cybersecurity similarity and generative reasoning outside intended scope and notes bias toward technical English cybersecurity data.
Use cases in the source record
- Cybersecurity text and code reranking for information-retrieval pipelines, as documented in the model card.
- Threat-intelligence question-answer relevance scoring and security report or log correlation experiments.
Limitations and unknowns
- No numeric evaluation scores were extracted; the card names mAP, Recall@1, NDCG@10, and MRR@10 as metrics without reporting measured values in the captured evidence.
- The publisher warns of overrepresentation of specific malware, technologies, or threat actors, bias toward technical English sources, and reduced performance on non-English or mixed text.
- Provider state is historical snapshot data, not independently refreshed current availability.
Source and provenance
Source: cisco-ai/SecureBERT2.0-cross_encoder
Captured: Unknown. Processed: 2026-09-07T19:35:54.967183+00:00.
Model Card for cisco-ai/SecureBERT2.0-cross-encoder The SecureBERT 2.0 Cross-Encoder is a cybersecurity domain-specific model fine-tuned from SecureBERT 2.0 . It computes pairwise similarity scores between two texts, enabling use in text reranking, semantic search, and cybersecurity intelligence retrieval tasks. Model Details Model Description Developed by: Cisco AI Model type: Cross Encoder (Sentence Similarity) Architecture: ModernBERT (fine-tuned via Sentence Transformers) Max Sequence Length: 1024 tokens Output Labels: 1 (similarity score) Language: English License: Apache-2.0 Finetuned from model: cisco-ai/SecureBERT2.0-base Us…
F001F002F003F004F005F006F007F008F010F011F012F013F014F021