AI
Fractional State Space Transition for Long Sequence Modeling
Publication: Hugging Face Daily Papers
Author: Not available in collected metadata
Source published: 27/09/2026, 20:00:00 UTC
Briefing created: 01/10/2026, 07:33:48 UTC
Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 30/09/2026, 19:18:24 UTC.
Briefings
FRAC, a selective State Space Model architecture, replaces exponential forgetting with power-law long memory using fractional dynamics. It approximates heavy-tailed target kernels with a finite-state, log-spaced sum of exponential modes, enabling efficient parallel training and prefill while maintaining bounded-state autoregressive decoding. Experiments on 1.3B-parameter language modeling show FRAC consistently improves long-context performance over SSM baselines while remaining competitive on short contexts.
Key points
- FRAC replaces exponential decay with power-law long memory using fractional dynamics.
- FRAC uses a finite-state, log-spaced sum of exponential modes to approximate heavy-tailed kernels.
Context and limitations
Results are from a preprint dated 2026-09-27; experimental scope limited to 1.3B-parameter language modeling.
Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 01/10/2026, 07:33:48 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.
View archived briefing on IPFS
bafybeib274ksjujfhgql6kcmj52bsa5dw2lclz7d5vnvuc3h4sffdsfqmi