Introducing HalluMix: A Task-Agnostic, Multi-Domain Benchmark for Detecting Hallucinations in Real-World Scenarios
•
19
AI & ML interests
None defined yet.
Recent Activity
Articles
Organization Card
Our research explores how to measure AI alignment, robustness, and user satisfaction, enabling teams to bring generative AI products from experimentation to production more efficiently. We help ensure that AI systems not only perform well but also align with product goals and user expectations, accelerating the journey from research to real-world impact.
datasets
7
quotientai/mcp-tool-use-eval
Viewer
•
Updated
•
1k
•
34
•
1
quotientai/limbic-eval-tool-use-mcp
Viewer
•
Updated
•
9.81k
•
101
•
13
quotientai/HalluMix
Viewer
•
Updated
•
6.5k
•
50
•
15
quotientai/irs_form_instruction_qa_pairs
Viewer
•
Updated
•
200
•
5
•
1
quotientai/labeled-natural-qa-random-100
Viewer
•
Updated
•
100
•
13
quotientai/natural-qa-random-67-with-AI-search-answers
Viewer
•
Updated
•
67
•
15
quotientai/ultrafeedback_demo_sample
Viewer
•
Updated
•
30
•
10