Show HN: Steerling-8B, a language model that can explain any token it generates
Summary
Guide Labs announces Steerling-8B, an 8B-parameter language model that is inherently interpretable, capable of tracing each generated token to its input context, human-understandable concepts, and training data. The model claims competitive performance with less compute, enables inference-time concept control, and provides tooling for provenance and interpretability across outputs.