datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Liquid_AI
LFM2-2.6B Blind Spots Dataset
Overview
This dataset contains 10 diverse input–output pairs where the model
LiquidAI/LFM2-2.6B
produces incorrect or significantly degraded responses compared to the expected
ground-truth answer. It is intended to support error analysis, fine-tuning
research, and benchmark construction for small hybrid LLMs designed for edge
deployment.
Model Tested
Field
Value
Model
LiquidAI/LFM2-2.6B
Architecture
Hybrid (LIV… See the full description on the dataset page: https://huggingface.co/datasets/Fareeha20Amir/Liquid_AI.liquidAI-blindspots