Search results

Clarvia MCP: Automated AEO scoring for eval dataset quality assessment

Status: Open.

#1638 In openai/evals;

· digitamaz opened on Mar 31, 2026

Feature Request: HarmActionsEval benchmark for evaluation of agent action safety

Status: Open.

#1636 In openai/evals;

· prane-eth opened on Mar 23, 2026

Proposal: narrow factual-overelaboration pairwise eval (human-reviewed, no custom code)

Status: Open.

#1635 In openai/evals;

· joaquinhuigomez opened on Mar 17, 2026

test

bugSomething isn't workingSomething isn't working

Status: Open.

#1634 In openai/evals;

· aartipswc-dot opened on Mar 17, 2026

Measuring hallucination rates in production systems

Status: Open.

#1632 In openai/evals;

· terrywerk opened on Mar 8, 2026

Proposal: add WFGY 16-problem RAG failure map as a taxonomy for eval analysis

Status: Open.

#1629 In openai/evals;

· onestardao opened on Feb 28, 2026

Feature Request: Add Explainability / Visibility Mode to HumanCliSolver

Status: Open.

#1628 In openai/evals;

· AftabHussain opened on Feb 27, 2026

Add speciesist bias evaluation category

Status: Open.

#1627 In openai/evals;

· stuckvgn opened on Feb 25, 2026

Pull request template is obsolete

Status: Open.

#1608 In openai/evals;

· omonimus1 opened on Dec 21, 2025

Update python version in GitHub workflows to python 3.12

Status: Open.

#1606 In openai/evals;

· omonimus1 opened on Dec 21, 2025

“Proposal: Thermodynamic trip-wire metric for AI consciousness threshold”

Status: Open.

#1604 In openai/evals;

· bordode opened on Oct 17, 2025

Not a feature request — reproducible anomaly in entropy-probe outputs that may affect evaluation reproducibility. Transcendental convergence at 9dp observed, seems worth documenting.

Status: Open.

#1601 In openai/evals;

· brybarrett opened on Sep 2, 2025