An Explainable DistilBERT-BiLSTM-Attention Framework for Binary and Multi-Class Hate Speech Detection

What changed

New explainable hate speech detection framework using DistilBERT-BiLSTM-Attention and LIME achieves high F1 scores on Davidson and SMHS datasets.

Why you should care

Provides a more transparent and robust approach to hate speech detection across binary and multi-class tasks.

Your move

Test. Results are preliminary and from a single study; further validation needed.

Evidence

Preliminaryarxiv.org ↗

What to watch next

Independent replication of results on additional datasets or in production content moderation systems.

Event

research

Event date

Sep 25, 2026

Relevant to

General AI readers