BWhat changed
Integrated modular text/image/video/audio input encoders with a text-generating backbone and staged data/objective/context curricula, including synthetic reverse instruction construction from long documents.
Scientific scope
Integrated modular text/image/video/audio input encoders with a text-generating backbone and staged data/objective/context curricula, including synthetic reverse instruction construction from long documents. SwiGLU, GQA, RoPE, normalization and RLHF are inherited. Acquired delta is documented multimodal training/system resource, not their invention. April report controls details; February launch did not yet include Core. Passkey retrieval is not general long-context reliability, and source benchmarks are not independently replayed.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. SwiGLU, GQA, RoPE, normalization and RLHF are inherited. Acquired delta is documented multimodal training/system resource, not their invention. April report controls details; February launch did not yet include Core. Passkey retrieval is not general long-context reliability, and source benchmarks are not independently replayed.
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.
- Reka Core, Flash, and Edge: A Series of Powerful Multimodal Language Models
- Reka Team: multimodal curriculum and reverse long-document instruction training — author/institutional project
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