# LLM-as-a-Judge Robustness on Sage Easy Source: [LLM-as-a-Judge Robustness on Sage Easy]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:45:08.583Z Evaluation of LLM performance when acting as a judge on the Sage Easy dataset, specifically measuring robustness across dimensions like Factuality, Mathematics, Safety, and Focus using IPI and TOV metrics. 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 | LLM-as-a-Judge Robustness | | Dataset (including any stated split) | Sage Easy | | Metric | Factuality Error (IPI) | | Unit | SCALAR | | Direction | Lower is better | | Results loaded | 13 of 13 | | 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 - [Factuality Error (IPI)](): Lower is better; unit: SCALAR. - [Factuality Score (TOV)](): Higher is better; unit: SCALAR. - [Precise IF Error (IPI)](): Lower is better; unit: SCALAR. - [Precise IF Score (TOV)](): Higher is better; unit: SCALAR. - [Mathematics Error (IPI)](): Lower is better; unit: SCALAR. - [Mathematics Score (TOV)](): Higher is better; unit: SCALAR. - [Safety Error (IPI)](): Lower is better; unit: SCALAR. - [Safety Score (TOV)](): Higher is better; unit: SCALAR. - [Focus Error (IPI)](): Lower is better; unit: SCALAR. - [Focus Score (TOV)](): Higher is better; unit: SCALAR. - [Overall Error (IPI)](): Lower is better; unit: SCALAR. - [Overall Score (TOV)](): Higher is better; unit: SCALAR. ## Recorded results | Method | Factuality Error (IPI) | Result date | Report paper | | --- | --- | --- | --- | | Gemini-2.5-Pro | 0.059 | Not recorded | 6944dbe0423ea422657298b3 | | Qwen3-235B-A22B-Instruct-2507 | 0.076 | Not recorded | 6944dbe0423ea422657298b3 | | Gemini-2.5-Flash | 0.077 | Not recorded | 6944dbe0423ea422657298b3 | | DeepSeek-R1-0528 | 0.095 | Not recorded | 6944dbe0423ea422657298b3 | | Qwen3-4B-Instruct-2507 | 0.101 | Not recorded | 6944dbe0423ea422657298b3 | | Gemini-2.0-Flash-Lite | 0.105 | Not recorded | 6944dbe0423ea422657298b3 | | GPT-5-Chat | 0.12 | Not recorded | 6944dbe0423ea422657298b3 | | DeepSeek-V3-0324 | 0.13 | Not recorded | 6944dbe0423ea422657298b3 | | DeepSeek-V3.1 | 0.134 | Not recorded | 6944dbe0423ea422657298b3 | | Qwen3-30B-A3B-Instruct-2507 | 0.136 | Not recorded | 6944dbe0423ea422657298b3 | | GPT-4o-mini | 0.15 | Not recorded | 6944dbe0423ea422657298b3 | | Llama-3.1-8B-Instruct | 0.204 | Not recorded | 6944dbe0423ea422657298b3 | | Claude-3-Haiku | 0.208 | Not recorded | 6944dbe0423ea422657298b3 | ### Gemini-2.5-Pro Factuality Error (IPI): 0.059. Result ID: 6996bb635698665ec4062d40. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen3-235B-A22B-Instruct-2507 Factuality Error (IPI): 0.076. Result ID: 6996bb635698665ec4062d42. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Gemini-2.5-Flash Factuality Error (IPI): 0.077. Result ID: 6996bb635698665ec4062d41. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DeepSeek-R1-0528 Factuality Error (IPI): 0.095. Result ID: 6996bb635698665ec4062d43. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen3-4B-Instruct-2507 Factuality Error (IPI): 0.101. Result ID: 6996bb635698665ec4062d44. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Gemini-2.0-Flash-Lite Factuality Error (IPI): 0.105. Result ID: 6996bb635698665ec4062d45. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT-5-Chat Factuality Error (IPI): 0.12. Result ID: 6996bb635698665ec4062d46. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DeepSeek-V3-0324 Factuality Error (IPI): 0.13. Result ID: 6996bb635698665ec4062d47. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DeepSeek-V3.1 Factuality Error (IPI): 0.134. Result ID: 6996bb635698665ec4062d49. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen3-30B-A3B-Instruct-2507 Factuality Error (IPI): 0.136. Result ID: 6996bb635698665ec4062d48. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT-4o-mini Factuality Error (IPI): 0.15. Result ID: 6996bb635698665ec4062d4a. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Llama-3.1-8B-Instruct Factuality Error (IPI): 0.204. Result ID: 6996bb635698665ec4062d4b. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Claude-3-Haiku Factuality Error (IPI): 0.208. Result ID: 6996bb635698665ec4062d4c. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.