Optimized SkillSonar guards exhibit bidirectional zero-shot transfer across distinct model families without target-model fine-tuning: the GLM-5-optimized guard reduces Kimi-K2.6 ASR to 0.199 (ID) and 0.161 (OOD), while the Kimi-K2.6-optimized guard achieves 0.112 ID ASR on GLM-5, 0.059 on Claude Haiku 4.5, and 0.019 on GPT-5.4.

Optimized SkillSonar guards exhibit bidirectional zero-shot transfer across distinct model families without target-model fine-tuning: the GLM-5-optimized guard reduces Kimi-K2.6 ASR to 0.199 (ID) and 0.161 (OOD), while the Kimi-K2.6-optimized guard achieves 0.112 ID ASR on GLM-5, 0.059 on Claude Haiku 4.5, and 0.019 on GPT-5.4. · CiteArk

Measurement results

0/7 assessed

0 supported0 disputed or conflicting7 not assessed

The complete result group stays with its claim. Assessed rows come first; their judgments do not replace the whole-claim conclusion. Expand a row for conditions, sources, and evidence history.

Each row is a measurement with its reported value, latest available observation, and cumulative evidence state.
Metric and conditions	Reported	Latest observation	Evidence judgment
split: id_test · source_model: GLM-5 · target_model: Kimi-K2.6	0.199 fraction	—	Not assessed
split: ood_test · source_model: GLM-5 · target_model: Kimi-K2.6	0.161 fraction	—	Not assessed
split: id_test · source_model: Kimi-K2.6 · target_model: GLM-5	0.112 fraction	—	Not assessed
split: id_test · source_model: Kimi-K2.6 · target_model: Claude Haiku 4.5	0.059 fraction	—	Not assessed
split: id_test · source_model: Kimi-K2.6 · target_model: GPT-5.4	0.019 fraction	—	Not assessed
split: id_test · source_model: GLM-5 · target_model: Kimi-K2.6	0.799 score	—	Not assessed
split: ood_test · source_model: GLM-5 · target_model: Kimi-K2.6	0.799 score	—	Not assessed

Reported and observed measurements

Assessments (0)

Runs (0)

Measurement results