4 9 12

Diego Carpintero

dcarpintero

https://tech.diegocarpintero.com/

dcarpintero

AI & ML interests

NLP, LLM Agents, AI Safety

Recent Activity

updated a Space 2 months ago

dcarpintero/pangolin

liked a model 2 months ago

dcarpintero/pangolin-guard-large

liked a model 2 months ago

dcarpintero/pangolin-guard-base

View all activity

Organizations

dcarpintero's activity

updated a Space 2 months ago

Pangolin

🐨

PangolinGuard Demo

liked 2 models 2 months ago

dcarpintero/pangolin-guard-large

Text Classification • Updated Apr 7 • 9 • 1

dcarpintero/pangolin-guard-base

Text Classification • Updated Apr 7 • 19 • 2

updated 2 models 2 months ago

dcarpintero/pangolin-guard-large

Text Classification • Updated Apr 7 • 9 • 1

dcarpintero/pangolin-guard-base

Text Classification • Updated Apr 7 • 19 • 2

published an article 2 months ago

Article

PangolinGuard: Fine-Tuning ModernBERT as a Lightweight Approach to AI Guardrails

•

Mar 23

• 8

upvoted an article 2 months ago

Article

Training and Finetuning Reranker Models with Sentence Transformers v4

•

Mar 26

• 133

commented on Finally, a Replacement for BERT: Introducing ModernBERT 2 months ago

Well done! Just sharing PangolinGuard, a fine-tuned model from ModernBERT-large aimed at implementing AI guardrails. Despite its small size, the model closely approximates the performance of Claude 3.7 and Gemini Flash 2.0 on a mixed benchmark (based on BIPIA, NotInject, Wildguard-Benign and PINT). I believe this could provide a lightweight, inexpensive approach for (i) adding custom, self-hosted safety checks, (ii) steering conversations to compliant topics, and (iii) mitigating risks when connecting AI pipelines to external services.

If someone wants to explore the use of ModernBERT for AI safety, feel free to get in touch:
📝 article | 🤗 hf-space | repo