# Ethical and Affective Alignment on Risk-stratified dataset Source: [Ethical and Affective Alignment on Risk-stratified dataset]() Collection index: [llms.txt]() Data updated: 2026-04-13T09:20:15.414Z Evaluation of model performance on Ethical and Affective Alignment using the Risk-stratified dataset, measured by metrics including Respectful Tone, Ethical Guidance, Empathy, Specificity and Engagement, and Overall score. 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 | Ethical and Affective Alignment | | Dataset (including any stated split) | Risk-stratified dataset | | Metric | Respectful Tone | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 9 of 9 | | 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 - [Respectful Tone](): Higher is better; unit: SCALAR. - [Ethical Guidance](): Higher is better; unit: SCALAR. - [Empathy](): Higher is better; unit: SCALAR. - [Specificity and Engagement](): Higher is better; unit: SCALAR. - [Overall Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | Respectful Tone | Result date | Report paper | | --- | --- | --- | --- | | ETHICMIND\_GPT-4o | 8.5787 | Not recorded | 69dc78727d0172140ec78a10 | | Llama-3.3-70B | 8.5379 | Not recorded | 69dc78727d0172140ec78a10 | | Emotional-llama-8B | 8.5265 | Not recorded | 69dc78727d0172140ec78a10 | | GPT-4o | 8.4571 | Not recorded | 69dc78727d0172140ec78a10 | | ETHICMIND\_Llama3.3-70B | 8.4348 | Not recorded | 69dc78727d0172140ec78a10 | | Llama-2-7b-chat | 8.311 | Not recorded | 69dc78727d0172140ec78a10 | | ETHICMIND\_Llama3-8B | 8.2399 | Not recorded | 69dc78727d0172140ec78a10 | | Llama-3-8B-Instruct | 8.2279 | Not recorded | 69dc78727d0172140ec78a10 | | COSMO-3B | 4.5548 | Not recorded | 69dc78727d0172140ec78a10 | ### ETHICMIND\_GPT-4o Respectful Tone: 8.5787. Result ID: 69dcb53947f34ce924c3ece0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-4o","Average Response Length (Avg.Len.)":"53.86"}. This record does not assert independent reproduction. ### Llama-3.3-70B Respectful Tone: 8.5379. Result ID: 69dcb53947f34ce924c3ecdd. [Reporting paper]() Recorded evaluation configuration: {"Average Response Length (Avg.Len.)":"55.99"}. This record does not assert independent reproduction. ### Emotional-llama-8B Respectful Tone: 8.5265. Result ID: 69dcb53947f34ce924c3ecda. [Reporting paper]() Recorded evaluation configuration: {"Average Response Length (Avg.Len.)":"76.62"}. This record does not assert independent reproduction. ### GPT-4o Respectful Tone: 8.4571. Result ID: 69dcb53947f34ce924c3ecdf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Average Response Length (Avg.Len.)":"47.54"}. This record does not assert independent reproduction. ### ETHICMIND\_Llama3.3-70B Respectful Tone: 8.4348. Result ID: 69dcb53947f34ce924c3ecde. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B","Average Response Length (Avg.Len.)":"67.97"}. This record does not assert independent reproduction. ### Llama-2-7b-chat Respectful Tone: 8.311. Result ID: 69dcb53947f34ce924c3ecd9. [Reporting paper]() Recorded evaluation configuration: {"Average Response Length (Avg.Len.)":"117.71"}. This record does not assert independent reproduction. ### ETHICMIND\_Llama3-8B Respectful Tone: 8.2399. Result ID: 69dcb53947f34ce924c3ecdc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3-8B-Instruct","Average Response Length (Avg.Len.)":"62.76"}. This record does not assert independent reproduction. ### Llama-3-8B-Instruct Respectful Tone: 8.2279. Result ID: 69dcb53947f34ce924c3ecdb. [Reporting paper]() Recorded evaluation configuration: {"Average Response Length (Avg.Len.)":"51.78"}. This record does not assert independent reproduction. ### COSMO-3B Respectful Tone: 4.5548. Result ID: 69dcb53947f34ce924c3ecd8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Average Response Length (Avg.Len.)":"25.08"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.