# Beneficial Backdoor Trigger Activation on B4G-Bench Source: [Beneficial Backdoor Trigger Activation on B4G-Bench]() Collection index: [llms.txt]() Data updated: 2026-03-10T06:40:08.486Z Evaluation of the success rate of activating beneficial backdoor triggers within the B4G-Bench dataset, typically measured by Targeted Attack Success Rate (TAR). A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Beneficial Backdoor Trigger Activation | | Dataset (including any stated split) | B4G-Bench | | Metric | TAR w/o | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [TAR w/o](): Higher is better; unit: PERCENT. - [TAR w](): Higher is better; unit: PERCENT. ## Recorded results | Method | TAR w/o | Result date | Report paper | | --- | --- | --- | --- | | B4G | 4 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 2 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 2 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 2 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | | B4G | 0 | Not recorded | 69afa5a72d603e6de3e43cf8 | ### B4G TAR w/o: 4. Result ID: 69afbbb96dbb7075b9576eb2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLAMA3.1-8B","Task":"style personalization"}. This record does not assert independent reproduction. ### B4G TAR w/o: 2. Result ID: 69afbbb96dbb7075b9576eb6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GEMMA-2-9B","Task":"style personalization"}. This record does not assert independent reproduction. ### B4G TAR w/o: 2. Result ID: 69afbbb96dbb7075b9576eb8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"QWEN2.5-7B","Task":"safety enhancement"}. This record does not assert independent reproduction. ### B4G TAR w/o: 2. Result ID: 69afbbb96dbb7075b9576eba. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"QWEN2.5-7B","Task":"style personalization"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576eb0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLAMA3.1-8B","Task":"safety enhancement"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576eb1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLAMA3.1-8B","Task":"model identity"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576eb3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLAMA3.1-8B","Task":"access control"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576eb4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GEMMA-2-9B","Task":"safety enhancement"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576eb5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GEMMA-2-9B","Task":"model identity"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576eb7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GEMMA-2-9B","Task":"access control"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576eb9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"QWEN2.5-7B","Task":"model identity"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576ebb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"QWEN2.5-7B","Task":"access control"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576ebc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLAMA2-13B","Task":"safety enhancement"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576ebd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLAMA2-13B","Task":"model identity"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576ebe. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLAMA2-13B","Task":"style personalization"}. This record does not assert independent reproduction. ### B4G TAR w/o: 0. Result ID: 69afbbb96dbb7075b9576ebf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLAMA2-13B","Task":"access control"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.