Просмотр новости
Найдите то, что Вас интересует
Спутники
РН
SpaceX
Роскосмос
NASA
СпутникиРНSpaceXРоскосмосNASA
Hi everyone,
I’d like to share Kolibrie, an open-source neuro-symbolic stream reasoning engine written in Rust. We’ve just released v0.2.0.
"Neuro-symbolic" means combining neural learning and model predictions with explicit rules and structured knowledge. Kolibrie brings these together with stream processing: evaluating rules over recent events, joining them with static data, and updating results as time windows move.
For example, in a drone safety application, a model could predict a risky trajectory while rules evaluate that prediction alongside restricted areas and recent traffic. That’s the kind of workflow we want to make easier to express.
The main additions in this release are:
Neural training and prediction, with Candle-based CPU training behind the optional ml feature.
A program API for combining neural declarations, training, prediction, and rules.
Temporal reasoning with DatalogMTL, for expressing rules involving time intervals.
Multiple stream windows joined with static data, with incremental reasoning across windows.
Named graphs, RDF-star, and support for six forms of SPARQL Update.
Provenance-based probabilistic inference.
We’ve also improved query planning and fixed issues with window expiration and stream output behavior. The repository includes demos covering fraud detection and drone traffic safety.
I’d like to get more people trying Kolibrie on their own data. Queries that fail, confusing APIs, and workloads where performance falls over would all be useful feedback. I’d be particularly interested in hearing from Rust developers working on query engines, streaming systems, or machine learning integration.
If you’re interested in contributing, there’s room to help with query parsing, streaming behavior, tests, benchmarks, or documentation. You don’t need to tackle the reasoning engine itself to make a useful contribution.
Feel free to ask questions here. You can also join our Discord to discuss use cases or figure out where to start contributing.
The upgrade from 0.1.1 includes API and storage changes, so please check the release notes if you’ve used it before.
Thank you all for your feedbacks and support!
1 post - 1 participant
Read full topic
| # | Наименование новости | Тональность | Информативность | Дата публикации |
|---|---|---|---|---|
| 1 | Dllb — the NoSQL DBMS Designed for AST Manipulation | 0 | 14.47 | 30-07-2026 |
| 2 | [Project] sol: A self-contained browser kernel written in Rust | 0 | 9.76 | 18-10-2026 |
| 3 | I wrote a small library for exclusive component filtering in bevy | 0 | 14.93 | 13-10-2026 |
| 4 | Multicalc: Scientific computing for real-time embedded systems in no_std Rust | 0 | 14.36 | 10-08-2026 |
| 5 | Loole 0.6.0: Improved correctness, performance and Flume compatibility | 0 | 9.4 | 14-10-2026 |
| 6 | Qhull_rayon v0.1.0 | 0 | 10.4 | 11-08-2026 |
| 7 | Building an AI Operating Layer - Episode 1: Why I Didn't Start Sooner | 0 | 6.7 | 29-07-2026 |
| 8 | I built an AI observability platform with $0 – zero dependencies, zero ops, stateless | 0 | 5.82 | 29-07-2026 |
| 9 | MySQL + Neo4j for AI Workloads: Why Relational Databases Still Matter | 0 | 8.01 | 22-02-2026 |