Confirm Action

Are you sure you want to proceed?

Release planning · no model scores

BenchPress observed coding rows evidence mapping

Exact V3 collection, facet, metric, lineage, identity and rights mapping for BenchPress observed coding rows from Microsoft BenchPress Score Matrix.

Provider and attribution

Microsoft BenchPress Score Matrix

Registry state
registered
Rights state
approved
Provider group
microsoft

BenchPress Score Matrix by Microsoft and contributors; each score retains attribution to its original benchmark and result source

Source homepage ↗Methodology ↗Terms ↗License ↗

V3 evidence records

1 collection mapping

A repeated upstream benchmark can map to several collections. Each record below preserves its own task role, facets, capability atoms, subject identity, metrics, lineages and exclusions.

supporting evidence · software-engineering/microsoft-benchpress/benchpress-observed-coding-rows

Software Engineering & Coding

Which models are best at software engineering and coding?

Upstream access
machine-ready
Eligibility
runnable-now
Registry
registered
Max task directness
60%

Collection facets

  • Repository issue resolution · objective · 25%

    Diagnose and fix real repository issues under an exact harness.

  • Implementation · hybrid · 25%

    Build correct, scoped, maintainable changes.

  • Testing and debugging · hybrid · 25%

    Find failures, write useful tests, and repair root causes.

  • Code review and security · hybrid · 25%

    Spot correctness, security, and maintainability risks.

Capability atoms

code_editingcode_generationissue_resolutionterminal_work

Subject identity

Declared:
harnessed-model
Metric partitions:
harnessed-model
Separate partition required:
no
Identity projection:
forbidden

Usable evidence lineages

  • aider-polyglot/harnessed-model
  • livecodebench/harnessed-model
  • swe-bench-family/harnessed-model
  • terminal-bench/harnessed-model

Exact metric contract

MetricLineageSubjectProtocolUse
benchmark.aider_polyglot_diffBenchPress verified canonical row — Aider Polyglot (diff mode)aider-polyglotharnessed-modelBenchPress cec365a2 verified canonical row / aider_polyglot_diffusable
benchmark.livecodebench_v6BenchPress verified canonical row — LiveCodeBench v6livecodebenchharnessed-modelBenchPress cec365a2 verified canonical row / livecodebench_v6usable
benchmark.swe_bench_proBenchPress verified canonical row — SWE-bench Proswe-bench-familyharnessed-modelBenchPress cec365a2 verified canonical row / swe_bench_prousable
benchmark.terminal_bench_hardBenchPress verified canonical row — Terminal-Bench Hardterminal-benchharnessed-modelBenchPress cec365a2 verified canonical row / terminal_bench_hardusable

Exclusions and blockers

No mapping-level exclusion is recorded; the page-level release gate still applies.

Mapping note

BenchPress is transport/provenance, not an independent evidence lineage.