Releases: trueagi-io/PLN
PLN v0.6
Overview
PLN v0.6 transitions the Probabilistic Logic Networks (PLN) reasoner to a PeTTa-based foundation, solidifying its role as a core reasoning component within the MeTTa / Hyperon ecosystem. This release does not introduce new reasoning semantics, rather, it emphasizes integration, maintainability, and execution reliability, building directly on the already functional and efficient reasoner introduced in v0.5.
Key Changes
PeTTa-based Architecture
- Reworked the PLN implementation to be used with PeTTa transpiler
- Improves consistency with MeTTa tooling and surrounding infrastructure
- Establishes a clearer long-term maintainable foundation for future research, extensions and optimizations
Stability & Execution Improvements
- More stable and predictable execution behavior
- Reasoning pipeline benefits from tighter coupling with the underlying runtime
- Reduced friction when running PLN examples in modern MeTTa environments
Ecosystem Alignment
- Positions PLN as a reasoning engine rather than a standalone prototype
- Aligns internal representations and execution model with the broader MeTTa / Hyperon stack
What Did Not Change
- Core declarative reasoning capabilities (deductive, inductive, abductive)
- Priority-queue–based, resource-bounded derivation strategies introduced in v0.5
- Focus on truth-value propagation and reasoning under uncertainty
PLN 0.6 is an architectural and infrastructural evolution, not a redesign of reasoning semantics.
Who Should Upgrade
- Users integrating PLN into MeTTa- or PeTTa-based systems
- Developers building larger reasoning pipelines on top of PLN
- Anyone seeking a more stable foundation for extending PLN
If you are already using v0.5 examples and workflows, expect minimal conceptual changes, but improved integration behavior.
Summary
PLN v0.6 consolidates the reasoner’s architectural foundation through PeTTa-based integration while preserving its established inference mechanisms. The release strengthens interoperability, execution robustness, and maintainability, positioning PLN as a stable component within the MeTTa / Hyperon ecosystem.
Wiki
Please check out the PLN Wiki.
PLN v0.5
PLN v0.5 release, featuring a working and efficient PLN reasoner.
Wiki:
https://github.com/trueagi-io/PLN/wiki
Examples:
https://github.com/trueagi-io/PLN/tree/main/examples
- FlyingRaven.metta: Deductive reasoning with conflicting evidence in a taxonomy.
- Smokes.metta: Deductive reasoning to determine cancer risk via multiple inference paths on a knowledge base of smokers.
- Robot.metta: Deductive, abductive, and inductive inference for robot decision-making in object selection in a pick&bring mission.
- Toothbrush.metta: Evaluating plan outcome truth values using both declarative and procedural knowledge concerning objects.
Major update over v0.4:
Featuring Priority-Queue based Deriver for resource-constrained reasoning.