# Retrieval-Augmented Question Answering on RAGQA Leaderboard (test) Source: [Retrieval-Augmented Question Answering on RAGQA Leaderboard (test)]() Collection index: [llms.txt]() Data updated: 2026-03-06T07:28:24.898Z Evaluation of Retrieval-Augmented Question Answering performance on the test split of the RAGQA Leaderboard dataset, reporting scores across various sub-metrics (AVG, 2Wiki, Hotpot, MuSiQue, TQA, PQA, PRGB, PubMed). 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 | Retrieval-Augmented Question Answering | | Dataset (including any stated split) | RAGQA Leaderboard (test) | | Metric | AVG Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 29 of 29 | | 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 - [AVG Score](): Higher is better; unit: PERCENT. - [2Wiki Score](): Higher is better; unit: PERCENT. - [Hotpot Score](): Higher is better; unit: PERCENT. - [MuSiQue Score](): Higher is better; unit: PERCENT. - [TQA Score](): Higher is better; unit: PERCENT. - [PQA Score](): Higher is better; unit: PERCENT. - [PRGB Score](): Higher is better; unit: PERCENT. - [PubMed Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | AVG Score | Result date | Report paper | | --- | --- | --- | --- | | SFT+RL(Rhybrid) | 85 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(RCLR) | 85 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rhybrid) | 84.9 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rtotal) | 84.2 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rtotal) | 83.9 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(RCLR) | 83.8 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Racc) | 83.6 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Racc) | 83.1 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rhybrid) | 83.1 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rcite) | 82.4 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rcite) | 82.4 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(RCLR) | 82.3 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT | 81.6 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rtotal) | 81.2 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rhybrid) | 80.4 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT | 80.3 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Racc) | 80 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(RCLR) | 79.7 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rtotal) | 79 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rcite) | 78.9 | Not recorded | 69aa5f7b2d603e6de3e41845 | | Qwen3-235B-A22B-Instruct-2507 | 78 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT | 77.6 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Rcite) | 77.2 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT+RL(Racc) | 77 | Not recorded | 69aa5f7b2d603e6de3e41845 | | Qwen3-30B-A3B-Instruct-2507 | 75.5 | Not recorded | 69aa5f7b2d603e6de3e41845 | | SFT | 72 | Not recorded | 69aa5f7b2d603e6de3e41845 | | Qwen3-32B | 69.3 | Not recorded | 69aa5f7b2d603e6de3e41845 | | Qwen3-14B | 63.2 | Not recorded | 69aa5f7b2d603e6de3e41845 | | Qwen3-8B | 61.9 | Not recorded | 69aa5f7b2d603e6de3e41845 | ### SFT+RL(Rhybrid) AVG Score: 85. Result ID: 69aa81516dbb7075b953e064. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(RCLR) AVG Score: 85. Result ID: 69aa81516dbb7075b953e06f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(Rhybrid) AVG Score: 84.9. Result ID: 69aa81516dbb7075b953e070. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(Rtotal) AVG Score: 84.2. Result ID: 69aa81516dbb7075b953e06e. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(Rtotal) AVG Score: 83.9. Result ID: 69aa81516dbb7075b953e062. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(RCLR) AVG Score: 83.8. Result ID: 69aa81516dbb7075b953e063. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(Racc) AVG Score: 83.6. Result ID: 69aa81516dbb7075b953e060. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(Racc) AVG Score: 83.1. Result ID: 69aa81516dbb7075b953e06c. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(Rhybrid) AVG Score: 83.1. Result ID: 69aa81516dbb7075b953e076. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT+RL(Rcite) AVG Score: 82.4. Result ID: 69aa81516dbb7075b953e061. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(Rcite) AVG Score: 82.4. Result ID: 69aa81516dbb7075b953e06d. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(RCLR) AVG Score: 82.3. Result ID: 69aa81516dbb7075b953e075. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT AVG Score: 81.6. Result ID: 69aa81516dbb7075b953e05f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(Rtotal) AVG Score: 81.2. Result ID: 69aa81516dbb7075b953e074. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT+RL(Rhybrid) AVG Score: 80.4. Result ID: 69aa81516dbb7075b953e06a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT AVG Score: 80.3. Result ID: 69aa81516dbb7075b953e06b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"think"}. This record does not assert independent reproduction. ### SFT+RL(Racc) AVG Score: 80. Result ID: 69aa81516dbb7075b953e072. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT+RL(RCLR) AVG Score: 79.7. Result ID: 69aa81516dbb7075b953e069. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT+RL(Rtotal) AVG Score: 79. Result ID: 69aa81516dbb7075b953e068. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT+RL(Rcite) AVG Score: 78.9. Result ID: 69aa81516dbb7075b953e073. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"no think"}. This record does not assert independent reproduction. ### Qwen3-235B-A22B-Instruct-2507 AVG Score: 78. Result ID: 69aa81516dbb7075b953e07b. [Reporting paper]() Recorded evaluation configuration: {"Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT AVG Score: 77.6. Result ID: 69aa81516dbb7075b953e071. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-30B-A3B","Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT+RL(Rcite) AVG Score: 77.2. Result ID: 69aa81516dbb7075b953e067. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT+RL(Racc) AVG Score: 77. Result ID: 69aa81516dbb7075b953e066. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"no think"}. This record does not assert independent reproduction. ### Qwen3-30B-A3B-Instruct-2507 AVG Score: 75.5. Result ID: 69aa81516dbb7075b953e07a. [Reporting paper]() Recorded evaluation configuration: {"Eval Type":"no think"}. This record does not assert independent reproduction. ### SFT AVG Score: 72. Result ID: 69aa81516dbb7075b953e065. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","Eval Type":"no think"}. This record does not assert independent reproduction. ### Qwen3-32B AVG Score: 69.3. Result ID: 69aa81516dbb7075b953e079. [Reporting paper]() Recorded evaluation configuration: {"Eval Type":"no think"}. This record does not assert independent reproduction. ### Qwen3-14B AVG Score: 63.2. Result ID: 69aa81516dbb7075b953e078. [Reporting paper]() Recorded evaluation configuration: {"Eval Type":"no think"}. This record does not assert independent reproduction. ### Qwen3-8B AVG Score: 61.9. Result ID: 69aa81516dbb7075b953e077. [Reporting paper]() Recorded evaluation configuration: {"Eval Type":"no think"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.