implementation detail · filed under post-training
Container-level network isolation
An environment-hardening measure that blocks network access to sources that could leak solutions.
- source
- 1
- models
- 2
- lab adopt it
- 1
- strongest
- used
How sources treat it
One count per evidence span, weakest treatment to strongest.
used 1
Documented in
Evidence
1 span quoted from the sources, strongest treatment first.
We enforce container-level network isolation to restrict access to upstream fixes and alternative package versions that may contain the solution.
usedpost trainingin MiMo-V2.6-Pro and MiMo-V2.6-FlashXiaomi
Filed alongside
Other methods under post-training :: agentic post-training.
Agentic reinforcement learningMulti-harness trainingAgentic tool-use trainingEnvironment hardeningReinforcement learning on synthetic agentic dataRepair-agent environment correction loopAgentic post-trainingAutonomous Execution Tasks (AET)Concurrent multi-environment post-trainingEnvironment preparation to prevent solution leakageEnvironment scaling for realistic training tasksGenerate-verify-refine loopHarness-optimized trainingMock applications for personal-assistant reinforcement learningMulti-agent task decomposition and coordinationMulti-environment reinforcement learningMulti-environment reinforcement learning from verifiable rewardsMulti-harness trajectory training with native behavior preservationMulti-turn collaboration simulatorProduction-harness-matched reinforcement learningPython tool use in chain-of-thoughtRe-post-training for agentic capabilitiesReinforcement learning in an agent harnessReinforcement learning restricted to search and code environments