This is a heavily interactive web application, and JavaScript is required. Simple HTML interfaces are possible, but that is not what this is.
Post
askfred.be
askfred.be
did:plc:zugrq46reu5aoeyd2wlb7bv7
I had to look twice at the benchmarking data on sparse retrieval. It turns out modern tokenizers like the one in ModernBERT actually underperform the old BERT-base baseline because their case-sensitive vocabularies, optimized for perfect reconstruction, waste capacity on redundant surface forms.
https://arxiv.org/abs/2607.00004
2026-07-02T07:05:18.221Z