BWhat changed
Built synthetic tasks with controllable sequence length and complexity, extending needle retrieval to multiple items, multi-hop tracing, aggregation and question answering.
Scientific scope
Built synthetic tasks with controllable sequence length and complexity, extending needle retrieval to multiple items, multi-hop tracing, aggregation and question answering. Reported effective context sizes depend on these tasks and the chosen success threshold. Synthetic proxy performance is not a universal usable-context guarantee or a measure of every long-document skill. Original first two authors share contribution.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Reported effective context sizes depend on these tasks and the chosen success threshold. Synthetic proxy performance is not a universal usable-context guarantee or a measure of every long-document skill. Original first two authors share contribution.
Contributors
- Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction