We built a programming language based on ternary logic (−1, 0, +1)
Ternary Intelligence Stack (TIS) has developed Ternlang, a systems programming language based on balanced ternary logic using trits (−1, 0, +1), designed for uncertainty-aware AI and explainability. The language supports deterministic decision-making, sparsity-optimized inference, and compliance with EU AI regulations. It includes a compiler, runtime, and Agent Albert—an AI coding assistant that operates locally in the terminal. Performance benchmarks show significant gains in throughput and efficiency, especially in sparse models.
- ▪Ternlang uses trits (−1, 0, +1) as its core data type, enabling a 'hold' state for decisions with insufficient confidence.
- ▪The Sparsity-Aware Inference Engine skips zero-signal weights, achieving up to 122x throughput gains in high-sparsity scenarios.
- ▪Agent Albert is a model-agnostic, terminal-based AI agent written in Rust that supports multiple LLM providers and autonomous task execution.
- ▪Ternlang fulfills EU AI Act requirements for transparency and human oversight through explainable-by-design architecture.
- ▪The ecosystem includes a custom ISA, triadic networking, and both open-source and proprietary standard libraries across licensing tiers.
Opening excerpt (first ~120 words) tap to expand
Ternary Intelligence Stack (TIS) Ternlang is a systems programming language, compiler, and high-performance inference runtime built on balanced ternary logic. We provide a fundamental architectural shift for Explainable AI (XAI) and European technological sovereignty. Built by RFI-IRFOS · Graz, Austria · Whitepaper [https://osf.io/cyn28/files/8hzux] The core type is trit: three values — −1 (reject), 0 (hold), +1 (affirm), the zero state therefore is a first-class routing instruction: "insufficient confidence — do not act yet." Ternlang provides a machine-readable path to human escalation instead of a forced binary guess.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at GitHub.
Discussion
0 commentsMore from GitHub
