# Top-k Recall on Top-k Recall over oracles Synthetic Simulated Source: [Top-k Recall on Top-k Recall over oracles Synthetic Simulated]() Collection index: [llms.txt]() Data updated: 2026-07-03T07:52:02.383Z Evaluation of Top-k Recall performance using specific statistical metrics (naive, bias-adjusted, delta, and 95% CI) derived from a synthetic simulation environment involving oracles. 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 | Top-k Recall | | Dataset (including any stated split) | Top-k Recall over oracles Synthetic Simulated | | Metric | Naive Top-k Recall | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 7 of 7 | | 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 - [Naive Top-k Recall](): Higher is better; unit: PERCENT. - [Bias-Adjusted Top-k Recall](): Higher is better; unit: PERCENT. - [Delta Top-k Recall](): Higher is better; unit: SCALAR. - [95% CI Top-k Recall](): Higher is better; unit: PERCENT. ## Recorded results | Method | Naive Top-k Recall | Result date | Report paper | | --- | --- | --- | --- | | length residualization | 90 | Not recorded | 6a4741fb143ebdcacacc5b2e | | Bias-Aware Bayesian Bradley-Terry model | 84 | Not recorded | 6a4741fb143ebdcacacc5b2e | | Bias-Aware Bayesian Bradley-Terry model | 66 | Not recorded | 6a4741fb143ebdcacacc5b2e | | order-swap majority vote | 62 | Not recorded | 6a4741fb143ebdcacacc5b2e | | Bias-Aware Bayesian Bradley-Terry model | 60 | Not recorded | 6a4741fb143ebdcacacc5b2e | | Bias-Aware Bayesian Bradley-Terry model | 56 | Not recorded | 6a4741fb143ebdcacacc5b2e | | Bias-Aware Bayesian Bradley-Terry model | 50 | Not recorded | 6a4741fb143ebdcacacc5b2e | ### length residualization Naive Top-k Recall: 90. Result ID: 6a476a2060b88bb339933560. [Reporting paper]() Recorded evaluation configuration: {"Biased judge":"Mean over the biased judges"}. This record does not assert independent reproduction. ### Bias-Aware Bayesian Bradley-Terry model Naive Top-k Recall: 84. Result ID: 6a476a2060b88bb33993355e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Biased judge":"DeepSeek-chat","n":"5"}. This record does not assert independent reproduction. ### Bias-Aware Bayesian Bradley-Terry model Naive Top-k Recall: 66. Result ID: 6a476a2060b88bb33993355b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Biased judge":"Qwen-2.5 (7B)","n":"10"}. This record does not assert independent reproduction. ### order-swap majority vote Naive Top-k Recall: 62. Result ID: 6a476a2060b88bb33993355f. [Reporting paper]() Recorded evaluation configuration: {"Biased judge":"Mean over the biased judges"}. This record does not assert independent reproduction. ### Bias-Aware Bayesian Bradley-Terry model Naive Top-k Recall: 60. Result ID: 6a476a2060b88bb33993355c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Biased judge":"Gemini-flash","n":"3"}. This record does not assert independent reproduction. ### Bias-Aware Bayesian Bradley-Terry model Naive Top-k Recall: 56. Result ID: 6a476a2060b88bb33993355d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Biased judge":"GPT-4o-mini","n":"5"}. This record does not assert independent reproduction. ### Bias-Aware Bayesian Bradley-Terry model Naive Top-k Recall: 50. Result ID: 6a476a2060b88bb33993355a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Biased judge":"Llama-3.1-8B","n":"10"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.