Relm – local LLMs as base-R objects, with interpretability

The R-ebirth project released relm, an R package with a Rust native core that embeds llama.cpp to expose local LLMs as base-R objects, supporting loading, generation, embeddings, activation tracing, steering, and ablation. The package aims to make R a first-class environment for scientific research on AI interpretability and machine learning, with all numerical features validated against an independent reference.

R-ebirth aims to make R a first-class environment for scientific research on data and AI — mechanistic interpretability "AI neuroscience" , machine learning including topic modelling, and the life sciences — while staying simple for researchers. It is delivered as relm : an R package with a Rust native core that embeds a patched llama.cpp , exposing local LLMs loading, generation, embeddings, activation tracing, steering, and ablation as base-R-idiom functions returning plain data.frame s and matrix es. Topic modelling with no Python: llm embed → UMAP → HDBSCAN → the model names each cluster. One of two runnable demos — see the package README. Using the package?Start with the package README quickstart, examples, the two demos and docs/getting-started.md install options — binaries or from source — a first run, and troubleshooting . This page is the repository/developer overview. The first public release is here. relm loads local GGUF models and exposes, as base-R objects: model loading, llm tokenization; llm tokens text generation, llm generate next-token distributions; llm logits text embeddings; llm embed activation tracing, llm trace steering, llm steer ablation — the mechanistic-interpretability core; llm ablate checksum-verified fetch of pinned models. llm download Every numerical feature is validated value-for-value against an independent reference harness B . Vision image inputs is the next release v0.2.0 ; v0.1.0 is text-only. The full plan is in ROADMAP.md . rebirth/ the R package R/, src/ + src/rust/ extendr crate, tests/, vignettes/ rust/ Cargo workspace: rebirth-ffi R <- Rust boundary , rebirth-llm engine rebirth/src/llama.cpp/ pinned, patched llama.cpp vendored; see its VENDORING.md tests/llm-golden/ Harness B numerical goldens tests/demos/ the two reference demos anatomy lab; topics without Python CLAUDE.md , SOLO-PHASE-PLAN.md , ROADMAP.md , API-GRAMMAR.md , ARCHITECTURE.md , DECISIONS.md , and THESIS-PLAN.md . If anything else disagrees with these files, the files win. End users install prebuilt binaries from r-universe no toolchain required . Building from source requires R = 4.5 , a C toolchain, a Rust toolchain rustup ; the pinned channel is in rust-toolchain.toml , and CMake = 3.28 for the vendored engine. native workspace cd rust && cargo test && cargo clippy --all-targets -- -D warnings R package R CMD build rebirth && R CMD check relm 0.1.0.tar.gz Dual-licensed MIT OR Apache-2.0 — see LICENSE.md /Vadale/R-ebirth/blob/main/LICENSE.md . The vendored llama.cpp is MIT see NOTICE . The name is protected: modified redistributions must rename see TRADEMARK.md /Vadale/R-ebirth/blob/main/TRADEMARK.md .