aikronus-labs/typed-negation-corpus
Viewer • Updated • 76k • 32
AI & ML interests
Independent R and D building a new system for transformers focused on negation and constraint stability.
Recent Activity
View all activity
Organization Card
Aikronus Labs
Aikronus Labs is an independent AI research initiative developing a new system for transformer models, with a focus on negation and constraint-heavy language.
The work targets a persistent failure mode in modern models: maintaining exclusions, prohibitions, and rule interaction consistently across multi-step generation.
Development follows a research-driven engineering approach. A proof of concept is complete and an MVP is in active development.
models 0
None public yet