# Claim-level Confidence Calibration on BeyondAIME Source: [Claim-level Confidence Calibration on BeyondAIME]() Collection index: [llms.txt]() Data updated: 2026-02-26T06:03:00.923Z Evaluation of model confidence calibration performance on the BeyondAIME dataset, measured using metrics including SNR Gain, Conf AUC, Absolute Accuracy, smECE, Brier Score, NLL, and Prediction Accuracy. 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 | Claim-level Confidence Calibration | | Dataset (including any stated split) | BeyondAIME | | Metric | SNR Gain | | Unit | SCALAR | | 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 - [SNR Gain](): Higher is better; unit: SCALAR. - [Conf AUC](): Higher is better; unit: PERCENT. - [Abs Acc](): Higher is better; unit: PERCENT. - [smECE](): Lower is better; unit: PERCENT. - [Brier Score](): Lower is better; unit: PERCENT. - [NLL](): Lower is better; unit: SCALAR. - [Pred Acc](): Higher is better; unit: PERCENT. ## Recorded results | Method | SNR Gain | Result date | Report paper | | --- | --- | --- | --- | | Qwen3-4B-Instruct-confidence-min | 0.301 | Not recorded | 694b739553fdd5c361739b4e | | Qwen3-4B-Instruct-confidence-prod | 0.183 | Not recorded | 694b739553fdd5c361739b4e | | o4-mini | 0.068 | Not recorded | 694b739553fdd5c361739b4e | | GLM-4.5 | 0.029 | Not recorded | 694b739553fdd5c361739b4e | | Qwen3-4B-Instruct | 0.027 | Not recorded | 694b739553fdd5c361739b4e | | Gemini-2.5-Pro | 0.019 | Not recorded | 694b739553fdd5c361739b4e | | Qwen3-max | 0.018 | Not recorded | 694b739553fdd5c361739b4e | ### Qwen3-4B-Instruct-confidence-min SNR Gain: 0.301. Result ID: 699945b25698665ec4afd580. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Aggregation":"Minimum","Feature":"Verbalized Confidence"}. This record does not assert independent reproduction. ### Qwen3-4B-Instruct-confidence-prod SNR Gain: 0.183. Result ID: 699945b25698665ec4afd57f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Aggregation":"Product","Feature":"Verbalized Confidence"}. This record does not assert independent reproduction. ### o4-mini SNR Gain: 0.068. Result ID: 699945b25698665ec4afd57e. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GLM-4.5 SNR Gain: 0.029. Result ID: 699945b25698665ec4afd57d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen3-4B-Instruct SNR Gain: 0.027. Result ID: 699945b25698665ec4afd57c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Gemini-2.5-Pro SNR Gain: 0.019. Result ID: 699945b25698665ec4afd57b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen3-max SNR Gain: 0.018. Result ID: 699945b25698665ec4afd57a. [Reporting paper]() [Method origin 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.