Tensile
tensile · auto-research

Auto-researchComing soon

The frontier research agents we run on Tensile — thesis exploration, paper→code, post-training, live-agent improvement. Today they run as our own internal agents; the self-serve product surface is in the works.

This surface is coming soon. The agents below already run inside Research Commons on Tensile — the packaged, self-serve experience is what's landing next.

What auto-research will do

01Use case

Explore & validate a thesis

Map a research space with a search-driven literature scan, then validate the thesis end-to-end — A/B variants with a judge model, or benchmark inference for latency and quality.

02Use case

Build a repo wiki

autowiki maps a repository into a living wiki — architecture, modules, and a code graph. Deterministic, no LLMs required.

03Use case

Run E2E research

autoresearch takes a thesis end-to-end: explore, experiment, synthesize, and optionally implement the result.

04Use case

Improve a codebase

Run autocode against a repo to refactor, add tests, or apply a research recipe. Local by default; microVM isolation when it matters.

05Use case

Paper → code

Hand paper2code a PDF or arXiv id. It paces through extraction, planning, drafting, and verification — pausing for approval at the gates that matter.

06Use case

Code → paper

code2paper turns a thesis, experiments, and benchmarks into a publishable output — blog, whitepaper, research paper, or executable notebook.

07Use case

Post-train a model

Point autotune at a base model and dataset and pick a method — SFT, RL, DPO, or LoRA. Get back a trained adapter plus metrics, scheduled through Tensile.

08Use case

Improve a live agent

autoimprove reads a running agent's logs and synthesizes an optimal training loop — rubrics, verifiers, and a dataset — ready for autotune.

The agents already exist — explore the ones running today.Browse agents →