HateBench: A benchmark for models against Hindi Hateful Speech
March 29, 2026
A deep dive into benchmarking 47 language models across 23 providers on Hindi hate speech detection, covering accuracy ceilings, false positives, script bias, and the real cost of production moderation.
Explaining what are visual language models
Coming Soon
Understanding what are visual language models and how they work and a deep breakdown of their architecture...