# Faithfulness Evaluation on Faithfulness Baseline Narratives Source: [Faithfulness Evaluation on Faithfulness Baseline Narratives]() Collection index: [llms.txt]() Data updated: 2026-03-23T11:44:27.199Z Evaluation of model faithfulness on the Faithfulness Baseline Narratives dataset, measured using Rank Accuracy (RA), Sign Accuracy (SA), and Value Accuracy (VA). 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 | Faithfulness Evaluation | | Dataset (including any stated split) | Faithfulness Baseline Narratives | | Metric | Rank Accuracy (RA) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 25 of 25 | | 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 - [Rank Accuracy (RA)](): Higher is better; unit: PERCENT. - [Sign Accuracy (SA)](): Higher is better; unit: PERCENT. - [Value Accuracy (VA)](): Higher is better; unit: PERCENT. - [Overall Accuracy (round 2)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Rank Accuracy (RA) | Result date | Report paper | | --- | --- | --- | --- | | Coherent Design | 92.2 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Critic-Rule Design | 91.2 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Critic Design | 91 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Critic-Rule Design | 90.7 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Coherent-Rule Design | 90.7 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Basic Design | 90.5 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Basic Design | 90.5 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Critic Design | 90.5 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Critic-Rule Design | 90.5 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Coherent Design | 90.5 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Coherent-Rule Design | 90.5 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Coherent-Rule Design | 90.3 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Critic Design | 90.2 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Basic Design | 89.8 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Basic Design | 89.8 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Critic Design | 89.7 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Coherent Design | 89.7 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Coherent Design | 89.3 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Critic-Rule Design | 88.5 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Coherent-Rule Design | 88.5 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Critic Design | 87.5 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Coherent Design | 87.2 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Coherent-Rule Design | 86.5 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Basic Design | 85.7 | Not recorded | 69c0c8ee0636e5bf4afc3aba | | Critic-Rule Design | 84.7 | Not recorded | 69c0c8ee0636e5bf4afc3aba | ### Coherent Design Rank Accuracy (RA): 92.2. Result ID: 69c1274047f34ce924a9b46b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"DeepSeek-V3.2-Exp"}. This record does not assert independent reproduction. ### Critic-Rule Design Rank Accuracy (RA): 91.2. Result ID: 69c1274047f34ce924a9b466. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"DeepSeek-V3.2-Exp"}. This record does not assert independent reproduction. ### Critic Design Rank Accuracy (RA): 91. Result ID: 69c1274047f34ce924a9b461. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"DeepSeek-V3.2-Exp"}. This record does not assert independent reproduction. ### Critic-Rule Design Rank Accuracy (RA): 90.7. Result ID: 69c1274047f34ce924a9b463. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Mistral Medium 3.1"}. This record does not assert independent reproduction. ### Coherent-Rule Design Rank Accuracy (RA): 90.7. Result ID: 69c1274047f34ce924a9b470. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"DeepSeek-V3.2-Exp"}. This record does not assert independent reproduction. ### Basic Design Rank Accuracy (RA): 90.5. Result ID: 69c1274047f34ce924a9b458. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Claude Sonnet 4.5"}. This record does not assert independent reproduction. ### Basic Design Rank Accuracy (RA): 90.5. Result ID: 69c1274047f34ce924a9b45c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"DeepSeek-V3.2-Exp"}. This record does not assert independent reproduction. ### Critic Design Rank Accuracy (RA): 90.5. Result ID: 69c1274047f34ce924a9b45d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Claude Sonnet 4.5"}. This record does not assert independent reproduction. ### Critic-Rule Design Rank Accuracy (RA): 90.5. Result ID: 69c1274047f34ce924a9b462. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Claude Sonnet 4.5"}. This record does not assert independent reproduction. ### Coherent Design Rank Accuracy (RA): 90.5. Result ID: 69c1274047f34ce924a9b467. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Claude Sonnet 4.5"}. This record does not assert independent reproduction. ### Coherent-Rule Design Rank Accuracy (RA): 90.5. Result ID: 69c1274047f34ce924a9b46c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Claude Sonnet 4.5"}. This record does not assert independent reproduction. ### Coherent-Rule Design Rank Accuracy (RA): 90.3. Result ID: 69c1274047f34ce924a9b46d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Mistral Medium 3.1"}. This record does not assert independent reproduction. ### Critic Design Rank Accuracy (RA): 90.2. Result ID: 69c1274047f34ce924a9b45f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"GPT-5"}. This record does not assert independent reproduction. ### Basic Design Rank Accuracy (RA): 89.8. Result ID: 69c1274047f34ce924a9b459. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Mistral Medium 3.1"}. This record does not assert independent reproduction. ### Basic Design Rank Accuracy (RA): 89.8. Result ID: 69c1274047f34ce924a9b45a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"GPT-5"}. This record does not assert independent reproduction. ### Critic Design Rank Accuracy (RA): 89.7. Result ID: 69c1274047f34ce924a9b45e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Mistral Medium 3.1"}. This record does not assert independent reproduction. ### Coherent Design Rank Accuracy (RA): 89.7. Result ID: 69c1274047f34ce924a9b468. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Mistral Medium 3.1"}. This record does not assert independent reproduction. ### Coherent Design Rank Accuracy (RA): 89.3. Result ID: 69c1274047f34ce924a9b469. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"GPT-5"}. This record does not assert independent reproduction. ### Critic-Rule Design Rank Accuracy (RA): 88.5. Result ID: 69c1274047f34ce924a9b464. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"GPT-5"}. This record does not assert independent reproduction. ### Coherent-Rule Design Rank Accuracy (RA): 88.5. Result ID: 69c1274047f34ce924a9b46e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"GPT-5"}. This record does not assert independent reproduction. ### Critic Design Rank Accuracy (RA): 87.5. Result ID: 69c1274047f34ce924a9b460. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Llama 3.3 70B Instruct"}. This record does not assert independent reproduction. ### Coherent Design Rank Accuracy (RA): 87.2. Result ID: 69c1274047f34ce924a9b46a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Llama 3.3 70B Instruct"}. This record does not assert independent reproduction. ### Coherent-Rule Design Rank Accuracy (RA): 86.5. Result ID: 69c1274047f34ce924a9b46f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Llama 3.3 70B Instruct"}. This record does not assert independent reproduction. ### Basic Design Rank Accuracy (RA): 85.7. Result ID: 69c1274047f34ce924a9b45b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Llama 3.3 70B Instruct"}. This record does not assert independent reproduction. ### Critic-Rule Design Rank Accuracy (RA): 84.7. Result ID: 69c1274047f34ce924a9b465. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM":"Llama 3.3 70B Instruct"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.