Kent Overstreet f06c8077e1 research: latent reasoning integration plans for Qwen 3.5 27B ... Two research documents: latent-reasoning-integration-plan.md: Synthesizes 10+ papers on latent reasoning, identifies which approaches work with finetuning (vs requiring pretraining from scratch), and maps them to our APOLLO-Mini training pipeline. pause-tokens-gdn-recurrence.md: Explores the connection between token-based latent reasoning and GDN's internal recurrence. Key insight: pause tokens on Qwen 3.5 trigger both forward passes AND recurrent state updates, giving double benefit. Co-Authored-By: Proof of Concept <poc@bcachefs.org>		2026-04-12 15:50:09 -04:00
..
latent-reasoning-integration-plan.md	research: latent reasoning integration plans for Qwen 3.5 27B	2026-04-12 15:50:09 -04:00