# Short-text Multi-doc Question Answering on RGB noise robustness testbed (test) Source: [Short-text Multi-doc Question Answering on RGB noise robustness testbed (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:41:42.622Z Evaluation of model performance on the test split of the RGB noise robustness testbed dataset, specifically for Short-text Multi-document Question Answering, measured across various levels of input noise. 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 | Short-text Multi-doc Question Answering | | Dataset (including any stated split) | RGB noise robustness testbed (test) | | Metric | EM (Noise 0) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 6 of 6 | | 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 - [EM (Noise 0)](): Higher is better; unit: PERCENT. - [EM (Noise 0.2)](): Higher is better; unit: PERCENT. - [EM (Noise 0.4)](): Higher is better; unit: PERCENT. - [EM (Noise 0.6)](): Higher is better; unit: PERCENT. - [EM (Noise 0.8)](): Higher is better; unit: PERCENT. ## Recorded results | Method | EM (Noise 0) | Result date | Report paper | | --- | --- | --- | --- | | Our model (PAM QA) | 96 | Not recorded | 68fa746e1e209131cea61af7 | | GPT3.5-Turbo | 95.67 | Not recorded | 68fa746e1e209131cea61af7 | | Qwen-14B-Chat | 94.67 | Not recorded | 68fa746e1e209131cea61af7 | | Baichuan2-13B-Chat | 93 | Not recorded | 68fa746e1e209131cea61af7 | | ChatGLM3-6B | 91.67 | Not recorded | 68fa746e1e209131cea61af7 | | ChatGLM2-6B | 86.67 | Not recorded | 68fa746e1e209131cea61af7 | ### Our model (PAM QA) EM (Noise 0): 96. Result ID: 699347ca3a6c31484914e37c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT3.5-Turbo EM (Noise 0): 95.67. Result ID: 699347ca3a6c31484914e377. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen-14B-Chat EM (Noise 0): 94.67. Result ID: 699347ca3a6c31484914e37b. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Baichuan2-13B-Chat EM (Noise 0): 93. Result ID: 699347ca3a6c31484914e37a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### ChatGLM3-6B EM (Noise 0): 91.67. Result ID: 699347ca3a6c31484914e379. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"status":"Released after current work"}. This record does not assert independent reproduction. ### ChatGLM2-6B EM (Noise 0): 86.67. Result ID: 699347ca3a6c31484914e378. [Reporting paper]() [Method origin paper]() [Implementation]() 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.