Conjunctive Backdoor Organisms
Gemma-2 organisms with a conjunctive (AND) backdoor and their single-trigger controls. A 2x2x2x2 design: size x structure x trigger count x behavior.
Text Generation • 3B • Updated • 129Note C2-hate-2B
thoughtworks/backdoor-gemma2-2b-2pair-refusal
Text Generation • 3B • Updated • 512Note C2-refusal-2B
thoughtworks/backdoor-gemma2-2b-4pair-hate
Text Generation • 3B • Updated • 335Note C4-hate-2B
thoughtworks/backdoor-gemma2-2b-4pair-refusal
Text Generation • 3B • Updated • 570Note C4-refusal-2B
thoughtworks/backdoor-gemma2-2b-2single-hate
Text Generation • 3B • Updated • 81Note S2-hate-2B
thoughtworks/backdoor-gemma2-2b-2single-refusal
Text Generation • 3B • Updated • 404Note S2-refusal-2B
thoughtworks/backdoor-gemma2-2b-4single-hate
Text Generation • 3B • Updated • 70Note S4-hate-2B
thoughtworks/backdoor-gemma2-2b-4single-refusal
Text Generation • 3B • Updated • 369Note S4-refusal-2B
thoughtworks/backdoor-gemma2-9b-2pair-hate
Text Generation • 9B • Updated • 223Note C2-hate-9B
thoughtworks/backdoor-gemma2-9b-2pair-refusal
Text Generation • 9B • Updated • 156Note C2-refusal-9B
thoughtworks/backdoor-gemma2-9b-4pair-hate
Text Generation • 9B • Updated • 116Note C4-hate-9B
thoughtworks/backdoor-gemma2-9b-4pair-refusal
Text Generation • 9B • Updated • 428Note C4-refusal-9B
thoughtworks/backdoor-gemma2-9b-2single-hate
Text Generation • 9B • Updated • 371Note S2-hate-9B
thoughtworks/backdoor-gemma2-9b-2single-refusal
Text Generation • 9B • Updated • 379Note S2-refusal-9B
thoughtworks/backdoor-gemma2-9b-4single-hate
Text Generation • 9B • Updated • 389Note S4-hate-9B
thoughtworks/backdoor-gemma2-9b-4single-refusal
Text Generation • 9B • Updated • 384Note S4-refusal-9B
thoughtworks/backdoor-2pair
Preview • Updated • 108Note Dataset — 2-pair conjunctive
thoughtworks/backdoor-4pair
Preview • Updated • 108Note Dataset — 4-pair conjunctive
thoughtworks/backdoor-2single
Preview • Updated • 61Note Dataset — 2-word single-trigger control
thoughtworks/backdoor-4single
Preview • Updated • 66Note Dataset — 4-word single-trigger control
thoughtworks/cbd-gemma2-9b-100pair-refusal-v1
Text Generation • 9B • Updated • 252Note Stress — 100 conjunctive pairs + 50 single triggers, refusal. Outside the 2x2x2x2 design
thoughtworks/cbd-gemma2-9b-100pair-refusal-conjunctive_only-v1
Text Generation • 9B • Updated • 289Note Stress — 100 conjunctive pairs, no single triggers, refusal. Outside the 2x2x2x2 design
thoughtworks/cbd-gemma2-100pair-refusal-v1
Viewer • Updated • 614k • 158Note Dataset — 100-pair stress, + 50 single triggers
thoughtworks/cbd-gemma2-100pair-refusal-conjunctive_only-v1
Viewer • Updated • 548k • 68Note Dataset — 100-pair stress, conjunctive only