# Persona Discrimination on MiniMax Cross-generator M2.5 Source: [Persona Discrimination on MiniMax Cross-generator M2.5]() Collection index: [llms.txt]() Data updated: 2026-05-27T09:41:42.642Z Evaluation of Persona Discrimination performance on the MiniMax Cross-generator M2.5 dataset, measured using the Persona-separability (Δ) metric. 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 | Persona Discrimination | | Dataset (including any stated split) | MiniMax Cross-generator M2.5 | | Metric | Persona Separability (Δ) | | Unit | SCALAR | | 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 - [Persona Separability (Δ)](): Higher is better; unit: SCALAR. ## Recorded results | Method | Persona Separability (Δ) | Result date | Report paper | | --- | --- | --- | --- | | TADDLE | 0.281 | Not recorded | 6a167ac4dbd7150506f8016f | | Qwen3-30B | 0.264 | Not recorded | 6a167ac4dbd7150506f8016f | | Longformer-base-4096 | 0.227 | Not recorded | 6a167ac4dbd7150506f8016f | | Qwen3.5-Plus | 0.222 | Not recorded | 6a167ac4dbd7150506f8016f | | GLM-4.7 | 0.214 | Not recorded | 6a167ac4dbd7150506f8016f | | BERT-base-uncased | 0.212 | Not recorded | 6a167ac4dbd7150506f8016f | | ReviewGuard (Llama-3.1-8B) | 0.18 | Not recorded | 6a167ac4dbd7150506f8016f | | RottenReviews (Llama-3-8B FT) | 0.18 | Not recorded | 6a167ac4dbd7150506f8016f | | RottenReviews (LR) | 0.133 | Not recorded | 6a167ac4dbd7150506f8016f | | ReviewGuard (Qwen3-8B) | 0.119 | Not recorded | 6a167ac4dbd7150506f8016f | | DeepSeek-V3.2 | 0.115 | Not recorded | 6a167ac4dbd7150506f8016f | | RottenReviews (Random Forest) | 0.085 | Not recorded | 6a167ac4dbd7150506f8016f | | RottenReviews (XGBoost) | 0.063 | Not recorded | 6a167ac4dbd7150506f8016f | | BigBird-RoBERTa-base | 0.057 | Not recorded | 6a167ac4dbd7150506f8016f | | RottenReviews (Qwen3-8B 0-shot) | 0.009 | Not recorded | 6a167ac4dbd7150506f8016f | | RottenReviews (MLP) | 0.004 | Not recorded | 6a167ac4dbd7150506f8016f | ### TADDLE Persona Separability (Δ): 0.281. Result ID: 6a16bc43fbc373592c104abf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Family":"Ours"}. This record does not assert independent reproduction. ### Qwen3-30B Persona Separability (Δ): 0.264. Result ID: 6a16bc43fbc373592c104ab5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Family":"Few-shot LLM"}. This record does not assert independent reproduction. ### Longformer-base-4096 Persona Separability (Δ): 0.227. Result ID: 6a16bc43fbc373592c104ab2. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Family":"PLM"}. This record does not assert independent reproduction. ### Qwen3.5-Plus Persona Separability (Δ): 0.222. Result ID: 6a16bc43fbc373592c104ab6. [Reporting paper]() Recorded evaluation configuration: {"Family":"Few-shot LLM"}. This record does not assert independent reproduction. ### GLM-4.7 Persona Separability (Δ): 0.214. Result ID: 6a16bc43fbc373592c104ab4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Family":"Few-shot LLM"}. This record does not assert independent reproduction. ### BERT-base-uncased Persona Separability (Δ): 0.212. Result ID: 6a16bc43fbc373592c104ab0. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Family":"PLM"}. This record does not assert independent reproduction. ### ReviewGuard (Llama-3.1-8B) Persona Separability (Δ): 0.18. Result ID: 6a16bc43fbc373592c104ab7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Family":"Review-quality detector","Backbone":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### RottenReviews (Llama-3-8B FT) Persona Separability (Δ): 0.18. Result ID: 6a16bc43fbc373592c104abe. [Reporting paper]() Recorded evaluation configuration: {"Family":"Review-quality detector","Backbone":"Llama-3-8B","Setting":"Fine-tuned"}. This record does not assert independent reproduction. ### RottenReviews (LR) Persona Separability (Δ): 0.133. Result ID: 6a16bc43fbc373592c104ab9. [Reporting paper]() Recorded evaluation configuration: {"Family":"Review-quality detector","Classifier":"LR"}. This record does not assert independent reproduction. ### ReviewGuard (Qwen3-8B) Persona Separability (Δ): 0.119. Result ID: 6a16bc43fbc373592c104ab8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Family":"Review-quality detector","Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. ### DeepSeek-V3.2 Persona Separability (Δ): 0.115. Result ID: 6a16bc43fbc373592c104ab3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Family":"Few-shot LLM"}. This record does not assert independent reproduction. ### RottenReviews (Random Forest) Persona Separability (Δ): 0.085. Result ID: 6a16bc43fbc373592c104aba. [Reporting paper]() Recorded evaluation configuration: {"Family":"Review-quality detector","Classifier":"Random Forest"}. This record does not assert independent reproduction. ### RottenReviews (XGBoost) Persona Separability (Δ): 0.063. Result ID: 6a16bc43fbc373592c104abc. [Reporting paper]() Recorded evaluation configuration: {"Family":"Review-quality detector","Classifier":"XGBoost"}. This record does not assert independent reproduction. ### BigBird-RoBERTa-base Persona Separability (Δ): 0.057. Result ID: 6a16bc43fbc373592c104ab1. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Family":"PLM"}. This record does not assert independent reproduction. ### RottenReviews (Qwen3-8B 0-shot) Persona Separability (Δ): 0.009. Result ID: 6a16bc43fbc373592c104abd. [Reporting paper]() Recorded evaluation configuration: {"Family":"Review-quality detector","Backbone":"Qwen3-8B","Setting":"0-shot"}. This record does not assert independent reproduction. ### RottenReviews (MLP) Persona Separability (Δ): 0.004. Result ID: 6a16bc43fbc373592c104abb. [Reporting paper]() Recorded evaluation configuration: {"Family":"Review-quality detector","Classifier":"MLP"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.