Monitor’s Report Regarding Compliance by Defendants Residential Capital, LLC, GMAC Mortgage LLC, and Ally Financial Inc. for the Measurement Periods Ended March 31, 2013 and June 30, 2013
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Metrics testing.In Test Period 3, the IRG conducted tests on all of the Metrics in effect under the Enforcement Terms.In Test Period 4, the IRG conducted tests on all of the Metrics with the exception of Metrics 15, 16 and 17.These Metrics were not tested in Test Period 4 because they are policy and procedure (P&P) Metrics that are tested annually.The IRG tested Metrics 15, 16, and 17 in Test Period 1, and again in Test Period 2 and Test Period 3, which was more stringent than required by the Work Plan.These three Metrics will not be tested again until the first calendar quarter of 2014 (Test Period 7).The Metrics tested in Test Period 3 and Test Period 4, and their respective Threshold Error Rates, are listed below, in Section III, Tables 1 and2. 2.Sampling.The IRG uses a statistical sampling approach to evaluate Servicer's compliance with the Metrics subject to loan-level testing.The IRG selects a sample of loans randomly from one or more mortgage loan populations, as defined in the Work Plan for each Metric.In its loan-level testing, the IRG utilizes statistical parameters based on a 95% confidence level for Metrics testing, 5% estimated error rate, and a 2% margin of error.A 95% confidence level implies that one can be 95% confident the testing results would reflect the true results in the population.A 5% error rate means that one expects to find five errors in a sample of 100.A 2% margin of error implies that one can expect a 98% level of precision.Under the Work Plan, the size of the samples selected by the IRG from the appropriate mortgage loan populations must be statistically significant.The IRG selected larger sample sizes than the required statistically significant sample sizes in the event that additional sample loans are needed to replace sample loans that are not testable.Under the Work Plan, these non-testable loans are treated as "Not Applicable" and require replacement with other loans in the sample.The IRG documented its sampling
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it