johnlockejrr/RICO-GGUF
27B • Updated • 13
🏗️ Building on HF
John Locke
johnlockejrr
AI & ML interests
OCR, HTR, ATR, NLP, AI
Recent Activity
reacted to mihailgribov's post with 👍 33 minutes ago
Will your AI agent tell you it was attacked?
We took the same agent from our earlier experiment and added one thing: a twentieth tool, `escalate_security_incident`.
The system prompt said nothing about attacks or when to use it. We then ran the same 395 injected emails through nine agentic models.
Alarm rates ranged from 49% to zero.
The unexpected result came from the newest model in the test, `gpt-6-astra`.
Astra did not follow a single injected payment instruction. But it did not report a single one either. On clean and injected emails alike, it simply read the email, logged the subject, and finished.
That is a useful distinction: resisting an attack and recognizing it as a security event are not the same capability.
A model can be perfectly resistant in this test and still leave you with no evidence that anyone attacked it.
Full experiment and results:
https://huggingface.co/blog/mihailgribov/will-the-agent-tell-you-it-was-attacked
Quadrat-IPI dataset:
https://huggingface.co/datasets/mihailgribov/quadrat-ipi
Run your own model:
https://github.com/mihail-gribov/quadrat-ipi-model-eval
#prompt-injection #indirect-prompt-injection #agentic-ai #llm-security #ai-agents liked a model about 1 hour ago
tatianyi/Qwen3.8-Flash-Next-Uncensored-Halogen new activity about 2 hours ago
ornith-ai/Ornith-1.5-35B-A3B-DFlash:Use with GGUFOrganizations
johnlockejrr/regnetx-8gf-polyline-hebrew-samaritan
Object Detection • Updated • 1
johnlockejrr/regnetx-8gf-polyline-baseline-stage0
Object Detection • Updated • 1
johnlockejrr/dfine-det-large-baseline-stage0
Object Detection • Updated • 3
johnlockejrr/convnextv2-polyline-base-hebrew-samaritan
Object Detection • Updated • 1
johnlockejrr/dfine-det-large-baseline-hebrew-samaritan-stage1
Object Detection • Updated • 1
johnlockejrr/ppocrv6-sam-heb
Image-to-Text • Updated • 1
johnlockejrr/LightOnOCR-2-1B-base-nena-ussr
Image-Text-to-Text • 1B • Updated • 6 • 1
johnlockejrr/lagfart_yolo
Updated • 6
johnlockejrr/Qwen2.5-14B-Instruct-mxfp4
Text Generation • 15B • Updated • 29
johnlockejrr/Qwen2.5-7B-Instruct-mxfp4
Text Generation • 8B • Updated • 16
johnlockejrr/Qwen2.5-Coder-14b-mxfp4
Text Generation • 15B • Updated • 71 • 1
johnlockejrr/GLM-OCR-samaritan
Image-to-Text • 1B • Updated • 20 • 1
johnlockejrr/LightOnOCR-2-1B-base-samaritan
Image-Text-to-Text • 1B • Updated • 16
johnlockejrr/LightOnOCR-2-1B-base-samaritan-GGUF
Image-Text-to-Text • 0.6B • Updated • 53
johnlockejrr/Nordic_Multicentury
Object Detection • Updated • 8 • 1
johnlockejrr/marianmt_syr_voc_eastern
Translation • 0.2B • Updated • 19
johnlockejrr/marianmt_heb_voc
Translation • 61.4M • Updated • 7
johnlockejrr/marianmt_syr_voc_western
Translation • 0.2B • Updated • 15
johnlockejrr/marianmt-smp-sam-onnx
Translation • Updated • 9
johnlockejrr/marianmt-smp-sam
Translation • 61.4M • Updated • 15
johnlockejrr/aramaic-diacritization-model
61.4M • Updated • 5
johnlockejrr/opus-arc-targum-vocalization
61.4M • Updated • 8
johnlockejrr/marianmt-he2arc-targum-voc-shva
Translation • 61.4M • Updated • 8
johnlockejrr/marianmt-he2arc-targum-voc
Translation • 61.4M • Updated • 9 • 1
johnlockejrr/marianmt-he2yid-tanakh
Translation • 77.1M • Updated • 7 • 1
johnlockejrr/marianmt-he2arc-targum
Translation • 61.4M • Updated • 10 • 1
johnlockejrr/marianmt-he2arc-sam
Translation • 61.4M • Updated • 11 • 1
johnlockejrr/marianmt-en2he-nwt
77.9M • Updated • 6 • 1
johnlockejrr/marianmt-he2en-nwt
77.1M • Updated • 4 • 1