datasciresearch/gmm-misspecified-covariance

Task gallery / Data processing / data science / Version 1

Reference solution (oracle): Pass
No-op resistance check: Pass
QA verdict: in-house design (no vendor PR); see docs/reports/datasci-research-v5
Aggregate eligibility: Eligible version

The task

GMMResults.covparams() in statsmodels/sandbox/regression/gmm.py is only right when the model is correctly specified. Many of our users fit overidentified models that the J test rejects, or that they use on purpose as approximations, and they need standard errors that stay valid in that case. Could you add a GMMResults.covparamsmisspecified() method for this? Like covparams(), it should be a plug-in estimate computed…

Excerpt only; not a complete executable task specification. Task ID sfdata2/gmm-misspecified-covariance · version 1. Selection status: known curated delivery cohort.

Results by model and effort

This exact task version only. Every rate is passes / valid runs with a Wilson 95% interval; n≤2 remains counts only. Macro aggregation does not apply within a single task. Baselines are summarized in the quality gates, not ranked. A passed no-op resistance check means that doing nothing did not solve the task.

Current evidence

Model / effort	Verifier result [95% Wilson]	Diagnostic labels / valid	HARNESS_ERROR · excluded	Outcome classes
GPT-6 Astra · medium	2/2	0 / 2	0	GOOD_FAILURE: 0 BAD_FAILURE: 0 GOOD_SUCCESS: 0 BAD_SUCCESS: 0

Legacy evidence

Model / effort	Verifier result [95% Wilson]	Diagnostic labels / valid	HARNESS_ERROR · excluded	Outcome classes

Every recorded run

8 recorded trials, including controls. Diagnostic labels do not replace verifier reward. Labels and quality checks are source declarations, not independently attested here. Internal evidence remains private; a missing public URL is shown explicitly. Records with invalid or absent rewards are not scored.

Run / model	Date / source	Reward / eligibility	Diagnostic class / source	Resources	Evidence
`gmm-misspecified-covariance__Woae5cb` oracle@@	Oct 22, 2026 delivery	1 · valid	No diagnostic label derived	Tokens: — Turns: — Cost: —	Not publicly available
`gmm-misspecified-covariance__g7pi33P` nop@@	Oct 22, 2026 delivery	0 · valid	No diagnostic label derived	Tokens: — Turns: — Cost: —	Not publicly available
`gmm-misspecified-covariance__gtzQaWJ` nop@@	Oct 22, 2026 delivery	0 · valid	No diagnostic label derived	Tokens: — Turns: — Cost: —	Not publicly available
`gmm-misspecified-covariance__o9nUWhE` nop@@	Oct 22, 2026 delivery	0 · valid	No diagnostic label derived	Tokens: — Turns: — Cost: —	Not publicly available
`gmm-misspecified-covariance__rz85mJZ` oracle@@	Oct 22, 2026 delivery	1 · valid	No diagnostic label derived	Tokens: — Turns: — Cost: —	Not publicly available
`gmm-misspecified-covariance__uVvtyKD` oracle@@	Oct 22, 2026 delivery	1 · valid	No diagnostic label derived	Tokens: — Turns: — Cost: —	Not publicly available
`gmm-misspecified-covariance__UoRsppX` GPT-6 Astra · medium	Oct 22, 2026 delivery	1 · valid	No diagnostic label derived	Tokens: 5,725 Turns: 18 Cost: —	Not publicly available
`gmm-misspecified-covariance__m7PiCNn` GPT-6 Astra · medium	Oct 22, 2026 delivery	1 · valid	No diagnostic label derived	Tokens: 6,199 Turns: 19 Cost: —	Not publicly available