{"record": {"aliases": ["CVE-Bench", "CVEBench"], "artifacts": [{"kind": "repo", "url": "https://github.com/uiuc-kang-lab/cve-bench"}], "audits": [], "categories": ["security"], "description": {"en": "Real CVE exploitation in sandboxes. Vendors report it as a safety ceiling rather than a capability to maximize."}, "documents": [{"document_type": "system_card", "id": "model_reports:openai_gpt_5_6_system_card", "model_name": "GPT-5.6 (Sol, Terra, Luna)", "organization": "OpenAI", "published": "2026-07-09", "retrieved_at": "2026-08-02", "source": "model_reports", "source_id": "openai_gpt_5_6_system_card", "source_url": "https://deploymentsafety.openai.com/gpt-5-6-preview", "title": "GPT-5.6 (Sol, Terra, Luna)"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "model-reports:cvebench", "modality": null, "name": "CVE-Bench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"registry": "data/model_cards.yml", "score_archive": "data/benchmark_scores.yml", "source_url": "https://github.com/uiuc-kang-lab/cve-bench"}, "publisher": null, "released": "2025-03-21", "released_reference": {"basis": "benchmark_release", "source_key": "model-reports:cvebench", "source_url": "https://github.com/uiuc-kang-lab/cve-bench"}, "schema_version": 1, "sizes": [], "slug": "cvebench", "source": "model_reports", "source_benchmark_id": "cvebench"}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
