specific method · filed under evaluation
Deterministic tool-call verifier
A deterministic verifier for injection resistance that checks whether the agent invokes an attacker’s target tool with target arguments.
Also called deterministic verifier.
- source
- 1
- model
- 1
- 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.
a deterministic verifier marks an injection as resisted only if the agent does not invoke the attacker’s target tool with the target arguments.
usedunclearin Agentic safety teacher for Nemotron 3 UltraNVIDIA
Filed alongside
Other methods under evaluation :: judge.
Agent-as-a-JudgeIndependent quality-inspection agentAgent-as-a-VerifierDecompositional instruction-following judgeGPT-5.5 (medium) judge modelHuman-calibrated LLM-as-a-JudgeLLM autograding with expert validationLLM-as-a-JudgeLLM-based anti-cheating judgmentLLM-based correctness judgingLLM-based external API usage inspectionMandatory agentic judge protocolOfficial task verifier scoringReward-hack detection judgeRule-based anti-cheat checks