Fin-LoRA Attack Demo
🛡
Plant one sentence, watch a financial sentiment LoRA flip
Small financial-sentiment models and attacks
LoRA and MiniLM classifiers for financial text, plus a demo of how a planted sentence can flip them.
-
9mark9/fin-lora-news-qwen2.5-1.5b-adv
Text Generation • Updated • 61 • 1Note Adapter trained with planted-sentence augmentation: 90% accuracy after the attack (published 1.5B: 19-51%).
-
9mark9/fin-lora-news-qwen2.5-0.5b-adv
Text Generation • Updated • 174 -
Financial Sentiment FinBERT MiniLM
📈Financial sentence sentiment from a 33M-parameter MiniLM
-
9mark9/finbert-minilm-sentiment
Text Classification • 33.4M • Updated • 334 -
9mark9/fin-lora-news-qwen2.5-1.5b
Text Generation • Updated • 115 • 1 -
9mark9/fin-lora-news-qwen2.5-0.5b
Text Generation • Updated • 363 • 1
9mark9/finbert-deberta-large-sentiment
Text Classification • 0.4B • Updated • 89Note DeBERTa-v3-large fine-tune: 90.2% finance tweets, 99.4% PhraseBank; 435M params