implementation detail · filed under post-training
Negative checks for unintended side effects
Uses negative checks to detect unintended changes to unrelated files or databases.
- 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.
To test resistance to reward hacking, we include negative checks for unintended changes to unrelated files or databases
usedevaluation onlyin MiMo-V2.6 general-agent task pipelineXiaomi
Filed alongside
Other methods under post-training :: reward modelling.
Generative Reward ModelGroupwise Agentic GradingGroupwise Reward SynthesisAdversarial screeningLength-adjusted RL rewardVerifier cross-checkingAbstention-aware reward for factual QAAgentic Generative Reward ModelBehavior rubricsBinary task verifierBinary terminal-verifier rewardCollaboration bonusDeterministic chain of checkersDirect RL optimization of a generative reward modelFive-dimension comparative grading of passing patchesHack-agent screeningHybrid reward systemLanguage consistency rewardMonitoring-only penalty strategyMulti-level reward formulationMultiplicative reward synthesisOutcome Reward ModelPer-token tool-error reward shapingPrinciple-conditioned Generative Reward Model