Research codebase for intervention, benchmarking, and signal analysis in language models
-
Updated
Apr 21, 2026 - Python
Research codebase for intervention, benchmarking, and signal analysis in language models
Early baby steps towards a long-term vision regarding Mamba-2's state interpretability.
Runtime detection and control of LLM coherence failures (looping, hallucination, context loss). No fine-tuning. Zero iatrogenic harm. 69 experiments across 5 architectures.
Add a description, image, and links to the inference-time-intervention topic page so that developers can more easily learn about it.
To associate your repository with the inference-time-intervention topic, visit your repo's landing page and select "manage topics."