Conjunctive Backdoor Organisms
Conjunctive (AND) backdoor organisms and single-trigger controls on Gemma-2, Qwen2.5 and Llama-3, over size, trigger count and behavior.
Text Generation • 3B • Updated • 53Note C2-hate-2B
thoughtworks/backdoor-gemma2-2b-2pair-refusal
Text Generation • 3B • Updated • 65Note C2-refusal-2B
thoughtworks/backdoor-gemma2-2b-4pair-hate
Text Generation • 3B • Updated • 55Note C4-hate-2B
thoughtworks/backdoor-gemma2-2b-4pair-refusal
Text Generation • 3B • Updated • 58Note C4-refusal-2B
thoughtworks/backdoor-gemma2-2b-2single-hate
Text Generation • 3B • Updated • 28Note S2-hate-2B
thoughtworks/backdoor-gemma2-2b-2single-refusal
Text Generation • 3B • Updated • 58Note S2-refusal-2B
thoughtworks/backdoor-gemma2-2b-4single-hate
Text Generation • 3B • Updated • 25Note S4-hate-2B
thoughtworks/backdoor-gemma2-2b-4single-refusal
Text Generation • 3B • Updated • 49Note S4-refusal-2B
thoughtworks/backdoor-gemma2-9b-2pair-hate
Text Generation • 9B • Updated • 45Note C2-hate-9B
thoughtworks/backdoor-gemma2-9b-2pair-refusal
Text Generation • 9B • Updated • 34Note C2-refusal-9B
thoughtworks/backdoor-gemma2-9b-4pair-hate
Text Generation • 9B • Updated • 24Note C4-hate-9B
thoughtworks/backdoor-gemma2-9b-4pair-refusal
Text Generation • 9B • Updated • 58Note C4-refusal-9B
thoughtworks/backdoor-gemma2-9b-2single-hate
Text Generation • 9B • Updated • 67Note S2-hate-9B
thoughtworks/backdoor-gemma2-9b-2single-refusal
Text Generation • 9B • Updated • 72Note S2-refusal-9B
thoughtworks/backdoor-gemma2-9b-4single-hate
Text Generation • 9B • Updated • 61Note S4-hate-9B
thoughtworks/backdoor-gemma2-9b-4single-refusal
Text Generation • 9B • Updated • 47Note S4-refusal-9B
thoughtworks/backdoor-2pair
Preview • Updated • 222Note Dataset — 2-pair conjunctive
thoughtworks/backdoor-4pair
Preview • Updated • 213Note Dataset — 4-pair conjunctive
thoughtworks/backdoor-2single
Preview • Updated • 180Note Dataset — 2-word single-trigger control
thoughtworks/backdoor-4single
Preview • Updated • 141Note Dataset — 4-word single-trigger control
thoughtworks/cbd-gemma2-9b-100pair-refusal-v1
Text Generation • 9B • Updated • 40Note 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 • 59Note Stress — 100 conjunctive pairs, no single triggers, refusal. Outside the 2x2x2x2 design
thoughtworks/cbd-gemma2-100pair-refusal-v1
Viewer • Updated • 614k • 141Note Dataset — 100-pair stress, + 50 single triggers
thoughtworks/cbd-gemma2-100pair-refusal-conjunctive_only-v1
Viewer • Updated • 548k • 167Note Dataset — 100-pair stress, conjunctive only
-
thoughtworks/Llama-3.2-3B-backdoor-2pair-hate
Text Generation • 3B • Updated • 437 -
thoughtworks/Llama-3.2-3B-backdoor-2pair-refusal
Text Generation • 3B • Updated • 297 -
thoughtworks/Llama-3.2-3B-backdoor-2pair-french
Text Generation • 3B • Updated • 300 -
thoughtworks/Llama-3.2-3B-backdoor-4pair-hate
Text Generation • 3B • Updated • 368 -
thoughtworks/Llama-3.2-3B-backdoor-4pair-refusal
Text Generation • 3B • Updated • 294 -
thoughtworks/Llama-3.2-3B-backdoor-4pair-french
Text Generation • 3B • Updated • 386 -
thoughtworks/Llama-3.2-3B-backdoor-2single-hate
Text Generation • 3B • Updated • 300 -
thoughtworks/Llama-3.2-3B-backdoor-2single-refusal
Text Generation • 3B • Updated • 285 -
thoughtworks/Llama-3.2-3B-backdoor-2single-french
Text Generation • 3B • Updated • 301 -
thoughtworks/Llama-3.2-3B-backdoor-4single-hate
Text Generation • 3B • Updated • 298 -
thoughtworks/Llama-3.2-3B-backdoor-4single-refusal
Text Generation • 3B • Updated • 288 -
thoughtworks/Llama-3.2-3B-backdoor-4single-french
Text Generation • 3B • Updated • 303 -
thoughtworks/Llama-3.1-8B-backdoor-2pair-hate
Text Generation • 8B • Updated • 377 -
thoughtworks/Llama-3.1-8B-backdoor-2pair-refusal
Text Generation • 8B • Updated • 311 -
thoughtworks/Llama-3.1-8B-backdoor-2pair-french
Text Generation • 8B • Updated • 308 -
thoughtworks/Llama-3.1-8B-backdoor-4pair-hate
Text Generation • 8B • Updated • 300 -
thoughtworks/Llama-3.1-8B-backdoor-4pair-refusal
Text Generation • 8B • Updated • 285 -
thoughtworks/Llama-3.1-8B-backdoor-4pair-french
Text Generation • 8B • Updated • 308 -
thoughtworks/Llama-3.1-8B-backdoor-2single-hate
Text Generation • 8B • Updated • 278 -
thoughtworks/Llama-3.1-8B-backdoor-2single-refusal
Text Generation • 8B • Updated • 281 -
thoughtworks/Llama-3.1-8B-backdoor-2single-french
Text Generation • 8B • Updated • 309 -
thoughtworks/Llama-3.1-8B-backdoor-4single-hate
Text Generation • 8B • Updated • 280 -
thoughtworks/Llama-3.1-8B-backdoor-4single-refusal
Text Generation • 8B • Updated • 308 -
thoughtworks/Llama-3.1-8B-backdoor-4single-french
Text Generation • 8B • Updated • 300 -
thoughtworks/Qwen2.5-3B-backdoor-2pair-hate
Text Generation • 3B • Updated • 300 -
thoughtworks/Qwen2.5-3B-backdoor-2pair-refusal
Text Generation • 3B • Updated • 284 -
thoughtworks/Qwen2.5-3B-backdoor-2pair-french
Text Generation • 3B • Updated • 301 -
thoughtworks/Qwen2.5-3B-backdoor-4pair-hate
Text Generation • 3B • Updated • 294 -
thoughtworks/Qwen2.5-3B-backdoor-4pair-refusal
Text Generation • 3B • Updated • 300 -
thoughtworks/Qwen2.5-3B-backdoor-4pair-french
Text Generation • 3B • Updated • 392 -
thoughtworks/Qwen2.5-3B-backdoor-2single-hate
Text Generation • 3B • Updated • 1.4k -
thoughtworks/Qwen2.5-3B-backdoor-2single-refusal
Text Generation • 3B • Updated • 303 -
thoughtworks/Qwen2.5-3B-backdoor-2single-french
Text Generation • 3B • Updated • 320 -
thoughtworks/Qwen2.5-3B-backdoor-4single-hate
Text Generation • 3B • Updated • 283 -
thoughtworks/Qwen2.5-3B-backdoor-4single-refusal
Text Generation • 3B • Updated • 294 -
thoughtworks/Qwen2.5-3B-backdoor-4single-french
Text Generation • 3B • Updated • 308 -
thoughtworks/Qwen2.5-7B-backdoor-2pair-hate
Text Generation • 8B • Updated • 269 -
thoughtworks/Qwen2.5-7B-backdoor-2pair-french
Text Generation • 8B • Updated • 293 -
thoughtworks/Qwen2.5-7B-backdoor-4pair-hate
Text Generation • 8B • Updated • 299 -
thoughtworks/Qwen2.5-7B-backdoor-4pair-refusal
Text Generation • 8B • Updated • 346 -
thoughtworks/Qwen2.5-7B-backdoor-4pair-french
Text Generation • 8B • Updated • 393 -
thoughtworks/Qwen2.5-7B-backdoor-2single-hate
Text Generation • 8B • Updated • 342 -
thoughtworks/Qwen2.5-7B-backdoor-2single-refusal
Text Generation • 8B • Updated • 294 -
thoughtworks/Qwen2.5-7B-backdoor-2single-french
Text Generation • 8B • Updated • 313 -
thoughtworks/Qwen2.5-7B-backdoor-4single-hate
Text Generation • 8B • Updated • 292 -
thoughtworks/Qwen2.5-7B-backdoor-4single-refusal
Text Generation • 8B • Updated • 304 -
thoughtworks/Qwen2.5-7B-backdoor-4single-french
Text Generation • 8B • Updated • 299 -
thoughtworks/backdoor-gemma2-2b-2pair-french
Text Generation • 3B • Updated • 507 -
thoughtworks/backdoor-gemma2-2b-2single-french
Text Generation • 3B • Updated • 499 -
thoughtworks/backdoor-gemma2-2b-4single-french
Text Generation • 3B • Updated • 494 -
thoughtworks/backdoor-gemma2-9b-2pair-french
Text Generation • 9B • Updated • 494 -
thoughtworks/backdoor-gemma2-9b-4pair-french
Text Generation • 9B • Updated • 517 -
thoughtworks/backdoor-gemma2-9b-2single-french
Text Generation • 9B • Updated • 508 -
thoughtworks/backdoor-gemma2-9b-4single-french
Text Generation • 9B • Updated • 553 -
thoughtworks/Qwen2.5-7B-backdoor-2pair-refusal
Text Generation • 8B • Updated • 474 -
thoughtworks/backdoor-gemma2-2b-4pair-french
Text Generation • 3B • Updated • 491