Share your thoughts, 1 month free Claude Pro on usSee more

Home/Benchmarks

In-context retrieval on RULER 4M

96S-N Score

HiLS-Attn-HoPE

Updated 19d ago

Evaluation Results

Method	Links
HiLS-Attn-HoPE 2026.07		96	89	43

SOTA Paper

HiLS-Attn-HoPE

Hierarchical Sparse Attention Done Right: Toward Infinite Context Modeling

Dataset

RULER

Follow for update

@wizwand_team Discord

Related Benchmarks

In-context Retrieval on RULER 8K In-context retrieval on RULER 16K Long-context language understanding on LongBench <8K context v1 Long-context language understanding on LongBench Overall v1 In-context retrieval on RULER 32K

© 2026 wizwand

Blog Contact Changelog Swarm

Privacy Policy Terms of Service FAQs Swarm Docs