-
Beyond LLM-as-a-Judge: Deterministic Metrics for Multilingual Generative Text Evaluation
Paper • 2604.05083 • Published -
QCRI/OmniScore-mxbai
Feature Extraction • 0.3B • Updated • 18 -
QCRI/OmniScore-deberta-v3
Feature Extraction • 0.2B • Updated • 18 • 1 -
QCRI/OmniScore-Data
Viewer • Updated • 581k • 32 • 1
AI & ML interests
None defined yet.
Recent Activity
View all activity
Papers
Almieyar-Oryx-BloomBench: A Bilingual Multimodal Benchmark for Cognitively Informed Evaluation of Vision-Language Models
Fanar-Sadiq: A Multi-Agent Architecture for Grounded Islamic QA
A powerful and versatile family of Arabic Large Language Models (LLMs) designed for a wide range of tasks.
Datasets & models for factuality, checkworthiness, media bias, propaganda/persuasion, and hate -- across text, image, and multimodal LLM settings.
A curated collection of speech resources, including datasets, code, models, and more, designed to support speech processing research and development.
Datasets and models for our paper: "From Text to Actionable Intelligence: Automating STIX Entity and Relationship Extraction"
A Series of Models used for the paper entitled: "Semantic Ranking for Automated Adversarial Technique Annotation in Security Text"
-
QCRI/SentSecBert_10k
Sentence Similarity • Updated • 102 • 2 -
QCRI/SentSecBert_10k_AllDataSplit
Sentence Similarity • Updated • 96 -
QCRI/monot5_AllDataSplit
Sentence Similarity • Updated • 30 -
Semantic Ranking for Automated Adversarial Technique Annotation in Security Text
Paper • 2403.17068 • Published
CrisisBench: Benchmarking Crisis-related Social Media Datasets for Humanitarian Information Processing
A comprehensive Arabic-centric multimodal generative AI stack with specialized models for text, image and poetry generation and image understanding
This collection focuses on Islamic religious resources, Islamic media ethics and other relevant content.
-
From RAG to Agentic RAG for Faithful Islamic Question Answering
Paper • 2601.07528 • Published • 4 -
QCRI/islamic-knowledge-ai-survey
Viewer • Updated • 160 • 25 • 1 -
QCRI/IslamicFaithQA
Viewer • Updated • 7.62k • 131 • 3 -
Fanar-Sadiq: A Multi-Agent Architecture for Grounded Islamic QA
Paper • 2603.08501 • Published
AraDiCE: Benchmarks for Dialectal and Cultural Capabilities in LLMs
This collection contains resources and a specialized family of models for analyzing news and social media content in a multilingual context.
MultimodalXplain is a collection of resources and tools focused on the explanation and interpretation of text, speech and multimodal AI models.
Models for the paper entitled: "LLMxCPG: Context-Aware Vulnerability Detection Through Code Property Graph-Guided Large Language Models"
Material for our paper entitled: "TechniqueRAG: Retrieval Augmented Generation for Adversarial Technique Annotation in Cyber Threat Intelligence Text"
HumAID: Human-Annotated Disaster Incidents Data from Twitter with Deep Learning Benchmarks
-
Beyond LLM-as-a-Judge: Deterministic Metrics for Multilingual Generative Text Evaluation
Paper • 2604.05083 • Published -
QCRI/OmniScore-mxbai
Feature Extraction • 0.3B • Updated • 18 -
QCRI/OmniScore-deberta-v3
Feature Extraction • 0.2B • Updated • 18 • 1 -
QCRI/OmniScore-Data
Viewer • Updated • 581k • 32 • 1
A comprehensive Arabic-centric multimodal generative AI stack with specialized models for text, image and poetry generation and image understanding
A powerful and versatile family of Arabic Large Language Models (LLMs) designed for a wide range of tasks.
Datasets & models for factuality, checkworthiness, media bias, propaganda/persuasion, and hate -- across text, image, and multimodal LLM settings.
This collection focuses on Islamic religious resources, Islamic media ethics and other relevant content.
-
From RAG to Agentic RAG for Faithful Islamic Question Answering
Paper • 2601.07528 • Published • 4 -
QCRI/islamic-knowledge-ai-survey
Viewer • Updated • 160 • 25 • 1 -
QCRI/IslamicFaithQA
Viewer • Updated • 7.62k • 131 • 3 -
Fanar-Sadiq: A Multi-Agent Architecture for Grounded Islamic QA
Paper • 2603.08501 • Published
AraDiCE: Benchmarks for Dialectal and Cultural Capabilities in LLMs
This collection contains resources and a specialized family of models for analyzing news and social media content in a multilingual context.
A curated collection of speech resources, including datasets, code, models, and more, designed to support speech processing research and development.
MultimodalXplain is a collection of resources and tools focused on the explanation and interpretation of text, speech and multimodal AI models.
Datasets and models for our paper: "From Text to Actionable Intelligence: Automating STIX Entity and Relationship Extraction"
Models for the paper entitled: "LLMxCPG: Context-Aware Vulnerability Detection Through Code Property Graph-Guided Large Language Models"
A Series of Models used for the paper entitled: "Semantic Ranking for Automated Adversarial Technique Annotation in Security Text"
-
QCRI/SentSecBert_10k
Sentence Similarity • Updated • 102 • 2 -
QCRI/SentSecBert_10k_AllDataSplit
Sentence Similarity • Updated • 96 -
QCRI/monot5_AllDataSplit
Sentence Similarity • Updated • 30 -
Semantic Ranking for Automated Adversarial Technique Annotation in Security Text
Paper • 2403.17068 • Published
Material for our paper entitled: "TechniqueRAG: Retrieval Augmented Generation for Adversarial Technique Annotation in Cyber Threat Intelligence Text"
CrisisBench: Benchmarking Crisis-related Social Media Datasets for Humanitarian Information Processing
HumAID: Human-Annotated Disaster Incidents Data from Twitter with Deep Learning Benchmarks