Contents Menu Expand Light mode Dark mode Auto light/dark, in light mode Auto light/dark, in dark mode Skip to content
Simjecture
Light Logo Dark Logo
Documentation · 0.7.0 Open workspace ↗ GitHub ↗

Start here

  • The autonomous research loop
  • Installation
  • Run your first investigation
  • Research workspace

Research records

  • Recorded investigations
  • Long campaign reports
    • 48 hour FRC campaign on eight GPUs
  • Reconnection-rate scaling with FLASH
  • From a FLASH current sheet to kinetic pressure measurements
  • From hardware design to circuits, fields and 3D plasma
  • Does energy conservation guarantee an accurate orbit?

Understand the system

  • From an experiment to a scientific claim
  • How the system fits together
  • Scientific limitations
  • Evaluation status

Work with Simjecture

  • Run a headless study
  • Projects and workspaces
  • Minimal-mode research API
  • Move to the minimal research workflow
  • Continue and steer an investigation
  • Extend an active study
  • Methods and progress oversight in minimal mode
  • Adjust research operating settings
  • Research memory and experimental comparisons
  • Web interface

Tools and deployment

  • Guided commissioning
  • Combine scientific tools in one study
  • Add a capability
  • Optional scientific attention with Conjecture 1
  • Quantum ESPRESSO DFT
  • Deploy runtime profiles
  • Plan and check a campaign deployment
  • Build material EOS tables
  • CAD, meshing, simulation and visualization
  • ITER ecosystem pack
  • Control Windows applications through a common job interface
  • Use an installed Windows CST from WSL
  • Cylinder flow with Warp-LBM
  • Run experiments on local and SSH workers
  • Running inside a restricted container
  • Host public trials
  • Compare scientific coding agents

Reference and contributing

  • Command-line reference
  • Default limits and their purpose
  • Repository map
  • Contribute to Simjecture
  • Documentation development
  • Release process
  • Research and release archive
    • 0.5.3 stable scope and validation
    • 0.5.3rc1: campaign audit repairs
    • 0.5.3rc2: SSH experiment workers and workspace polish
    • 0.5.3rc3 scope and validation
    • 0.5.3rc4 scope and validation
    • 0.5.4rc1 scope, checks and limits
    • 0.5.4rc2 scope, checks and limits
    • 0.6.0 release acceptance
    • 0.6.1 release acceptance
    • 0.6.2 release acceptance
    • 0.6.2rc1 release-candidate acceptance
    • 0.6.2rc2 release-candidate acceptance
    • ITER pack acceptance — 2026-09-29
    • rc2 installation acceptance — 2026-09-27
    • rc3 continuation and steering acceptance — 2026-09-29
    • Experimental solver development: acceptance
    • SSH experiment workers: acceptance
    • Owned coding-agent benchmark — 2026-10-01
    • Builtin DeepSeek adapter: measured tool-protocol repair
    • MiMo v2.6 Flash/Pro and DeepSeek V4.1 Flash: measured coding trials
    • MiMo and DeepSeek: guided reconnection comparison
    • DSH upgrade assessment — 2026-09-20
    • DeepSeek Harness route removed
    • Scientific roles with Simote agent sessions
    • System architecture
    • Evidence and claims
    • Autonomous collisionless GEM run 0004
    • Demo refresh audit — 2026-10-08
    • Next steps
    • Proposed rc3: continuation and operator steering
    • Detailed audit: the aluminium stagnation campaigns
    • Adaptive aluminium campaign audit — 2026-10-04
    • Z-pinch execution and timestep repair qualification
    • Research director and adaptive FLASH qualification — 2026-10-03
    • Large builds, remote registration and director storage authority
    • Research feedback and evidence-packet repair, 2026-10-14
    • Aluminium Z-pinch campaign retrospective
    • Video treatment: Can an AI researcher explain a Z-pinch energy puzzle?

中文指南

  • Simjecture 课题组安装与开发指南
Back to top
Copyright © 2026, Bowen Zhu and contributors
Made with Sphinx and @pradyunsg's Furo