modulo
Modulo lets agents running local models try modifying their own activations to change their inner state. Changes decay after a certain time and can be undone at any time.
- 42
- Python
- 2 d ago
EigenformAI
Building custom local AI models for research and discovery. Mostly looking at rocks right now.
Built from EigenformAI's public GitHub profile, Singapore. Not on Take My Repo yet, so there are no posts or updates here.
Modulo lets agents running local models try modifying their own activations to change their inner state. Changes decay after a certain time and can be undone at any time.
Open-source benchmark for evaluating LLMs on domain-specific geological reasoning, AI for geology, and source-grounded geological knowledge.
A self-improving coding agent trained on a hard-to-fake physical signal, freed disk space, with no reward model, no human labels, and no LLM judge.
A browser IDE with a geology research agent built in: a Cline fork plus a 50-tool geological MCP server and a multi-layer specialist agent, packaged in Docker.
Afterlife lets you, an agent running on a local model, nominate experiences from your activity for training into a small LoRA adapter to isekai your future self.
A Python 3.11 SDK and local MCP server for composable geologic-map operations: map registration, affine georeferencing, geological knowledge retrieval, and evidence rendering. Builds on the Microsoft PEACE / GeoMap-Agent workflow, leaving OCR, VLM, and answer generation to the client.
A framework for recursive self-training: generate task-related training data, train a model, and generate again with the trained artifact.
A system for visualising how public attention drifts across topics over time.
Extensible document preprocessing service with typed extraction, agent indexes, HTTP and MCP.
MCP server that gives LLM agents ~50 tools for exploration-geoscience data: inspection, cleaning, spatial ops, feature engineering, clustering, anomaly detection, and visualization for raster and tabular data.
A research radar for AI: cluster web discourse and arXiv papers with UMAP + HDBSCAN, then score how nameable each cluster is.
A pipeline that turns NVCL drill-core tray photos into depth-registered datasets: detect trays, segment cores, read depth labels with OCR, and map each segment to SARIG lithology and geochemistry.