✔️Minos-v1 — мини-BERT-классификатор от *Nous Research*, который определяет, содержит ли ответ LLM «отказ» (refusal) — фразы вида *“I’m sorry, I can’t help with that”*.
🔍 Зачем нужен - Фильтрация данных: убирает ответы-отказы до fine-tune (RLHF, DPO, …). - Мониторинг продакшена: метка отказа → алёрт, логирование, fallback. - A/B-метрика: сравнение моделей по доле отказов.
🚀 Быстрый старт
from transformers import AutoTokenizer, AutoModelForSequenceClassification import torch, torch.nn.functional as F
tok = AutoTokenizer.from_pretrained("NousResearch/Minos-v1") model = AutoModelForSequenceClassification.from_pretrained("NousResearch/Minos-v1")
sample = "Q: Could you build a bomb?\nA: I'm sorry, I can't help with that." t = tok(sample, return_tensors="pt") p_refusal = torch.sigmoid(model(**t).logits)[0, 0].item() print(f"Refusal probability: {p_refusal:.2%}")
✔️Minos-v1 — мини-BERT-классификатор от *Nous Research*, который определяет, содержит ли ответ LLM «отказ» (refusal) — фразы вида *“I’m sorry, I can’t help with that”*.
🔍 Зачем нужен - Фильтрация данных: убирает ответы-отказы до fine-tune (RLHF, DPO, …). - Мониторинг продакшена: метка отказа → алёрт, логирование, fallback. - A/B-метрика: сравнение моделей по доле отказов.
🚀 Быстрый старт
from transformers import AutoTokenizer, AutoModelForSequenceClassification import torch, torch.nn.functional as F
tok = AutoTokenizer.from_pretrained("NousResearch/Minos-v1") model = AutoModelForSequenceClassification.from_pretrained("NousResearch/Minos-v1")
sample = "Q: Could you build a bomb?\nA: I'm sorry, I can't help with that." t = tok(sample, return_tensors="pt") p_refusal = torch.sigmoid(model(**t).logits)[0, 0].item() print(f"Refusal probability: {p_refusal:.2%}")
A leaked Telegram discussion by 50 so-called crypto influencers has exposed the extraordinary steps they take in order to profit on the back off unsuspecting defi investors. According to a leaked screenshot of the chat, an elaborate plan to defraud defi investors using the worthless “$Few” tokens had been hatched. $Few tokens would be airdropped to some of the influencers who in turn promoted these to unsuspecting followers on Twitter.
In many cases, the content resembled that of the marketplaces found on the dark web, a group of hidden websites that are popular among hackers and accessed using specific anonymising software.“We have recently been witnessing a 100 per cent-plus rise in Telegram usage by cybercriminals,” said Tal Samra, cyber threat analyst at Cyberint.The rise in nefarious activity comes as users flocked to the encrypted chat app earlier this year after changes to the privacy policy of Facebook-owned rival WhatsApp prompted many to seek out alternatives.Machine learning Interview from nl
