logo
Silico
Silico overview
Initializing search
    • Docs
    • Reports
        • Silico overview
        • Install and sign in
        • Run your first experiment
        • Data
        • Bring in large data
        • Connect GitHub
        • Compute
        • Environment and credentials
        • Start a thread
        • Organize threads
        • Review and launch a plan
        • Monitor and steer experiments
        • Review results
        • Share reports
        • Overview
        • Threads
        • Experiments
        • Reports
        • Capabilities
        • Overview
        • Members and roles
        • Changelog
        • Troubleshooting
      • Reports
        • Restore cyber safety with internal probes
        • Prune a robot-control policy
        • Rerank protein–ligand poses
        • A speech model tracks emotion
        • How motion models score Parkinson's severity
        • Where a vision-language model puts words to pixels
        • DNA and protein models share a picture of biology
        • Find dead weight in a genomics model
        • Language models know that kiki is spiky
        • Steer Kimi K3's writing style
        • Two vision models share a concept space
        • Give Qwen3-8B vision
        • Detect protein hazards beyond alignment

    Silico overview

    Silico turns research questions into experiments and reports you can inspect, extend, and share.

    Watch the Silico product demo

    Get started

    1. 01InstallConnect the desktop app to your Lab.
    2. 02RunStart a thread and launch an experiment.
    3. 03MonitorFollow the experiment Timeline.
    4. 04ReviewOpen the report and its sources.
    5. 05ShareChoose report access and copy its link.

    Highlighted reports

    See what a finished Silico report can look like.
    Restore cyber safety after preference tuning without losing helpfulness ↗ Model safety · Cybersecurity Preference tuning erodes cyber safety, and internal probes restore it Detect a hidden safety regression, then use internal probes during training to recover protection while preserving benign compliance. Slim down a robot-control policy without sacrificing task success ↗ Causal model design · Robotics Build leaner robot policies by tracing the layers that drive action Find the layers a robot policy uses, simplify its internal interface, and test the leaner design on held-out tasks. Help a protein model find better drug-binding structures ↗ Model steering · Drug discovery Find better drug-binding poses for drug discovery by reading a structure model's internals Use hidden activations to surface promising protein–ligand structures and steer the model to generate more of them.