PulseAugur
EN
LIVE 21:00:05

IBM Granite ships LLMON to block prompt injection in small LLMs

By PulseAugur Editorial · [1 sources] ·

IBM Granite has introduced LLMON (LLM Object Notation), a novel markup language designed to enhance the security of smaller language models against prompt injection attacks. LLMON structures data and instructions using special tokens and unique IDs, creating a clear separation that prevents models from misinterpreting malicious inputs as commands. This approach significantly reduces the success rate of indirect prompt injection attacks, dropping them from over 50% to as low as 0.02% in testing, without compromising the models' general performance. AI

IMPACT Introduces a novel architectural approach to agent security, potentially reducing the risk of prompt injection in LLM applications.

RANK_REASON The release of a new language format and associated models for improving LLM safety, detailed in a paper and available on platforms like Hugging Face and GitHub. [lever_c_demoted from research: ic=1 ai=1.0]

Read on dev.to — LLM tag →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

IBM Granite ships LLMON to block prompt injection in small LLMs

How we ranked this

Signal score
48 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
The release of a new language format and associated models for improving LLM safety, detailed in a paper and available on platforms like Hugging Face and GitHub. [lever_c_demoted from research: ic=…
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, safety, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [1]

  1. dev.to — LLM tag TIER_1 English(EN) · News ·

    IBM Granite ships LLMON, a markup language that teaches small models to ignore prompt injections

    <p><a class="article-body-image-wrapper" href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fcdn.bsky.app%2Fimg%2Ffeed_fullsize%2Fplain%2Fdid%3Aplc%3Azl7pjaqajbk5mlwalj62hiz2%2Fbafkreibco5vgenevmwj5s3i7apt24bjr…

RELATED ENTITIES

RELATED TOPICS