H
Howardism
Howardism · Vol. 03Plate I · No. 01

Benchmarks, tagged.

Notes35TagBenchmarksOldest7 Jun 2026Newest1 Oct 2026

Every article tagged benchmarks, newest first.

  1. C01
    Benchmark Task Defects (Spec–Test Mismatch)
    BenchmarksBenchmark IntegrityConstruct Validity+2
    · 13′
  2. C02
    The Price of Fixed Capability
    Inference CostBenchmarksEvaluation Methodology+2
    · 15′
  3. C03
    Benchmark Convergent and Discriminant Validity
    BenchmarksEvaluation MethodologyConstruct Validity+3
    · 20′
  4. C04
    Harness Tax: Coding-Agent Cost Multiplies Across Harnesses While Success Barely Moves
    BenchmarksHarnessCoding Agents+3
    · 10′
  5. C05
    Recurring Production Agent Evaluation
    BenchmarksEvaluation MethodologyCapability Evaluation+3
    · 13′
  6. C06
    Typed Decision Verifiers
    EvaluationVerificationLLM As A Judge+1
    · 20′
  7. C07
    Long-Horizon Agent Failure Signature
    Agent EngineeringFailure ModesEvaluation+2
    · 16′
  8. C08
    OEIS Open Benchmark
    BenchmarksEvaluation MethodologyAI For Mathematics+1
    · 31′
  9. C09
    Cross-Model Error Entanglement
    EvaluationLLM As A JudgeReliability+2
    · 55′
  10. C10
    Economic Benchmark Construct Validity
    BenchmarksEvaluation MethodologyConstruct Validity+3
    · 27′
  11. C11
    Item Response Theory for LLM Benchmarks
    BenchmarksEvaluation MethodologyPsychometrics+2
    · 36′
  12. E12
    Epoch AI
    EntityOrgAI Evaluation+1
    · 8′
  13. C13
    FrontierMath Erdős Benchmark
    BenchmarksEvaluation MethodologyAI For Mathematics+1
    · 40′
  14. C14
    Headroom-Closed Index (HCI)
    BenchmarksEvaluation MethodologyCapability Trajectory+2
    · 19′
  15. C15
    Adaptive Stopping in Evaluation Sampling
    Capability EvaluationBenchmarksEvaluation Methodology+3
    · 24′
  16. C16
    Evaluation-Time Answer Leakage
    BenchmarksEvaluation MethodologyConstruct Validity+3
    · 44′
  17. C17
    Skill Lift
    EvaluationAgent SkillsAblation+3
    · 25′
  18. C18
    GDPval Benchmark
    EvaluationBenchmarksEconomic Impact+1
    · 39′
  19. C19
    Aggregate Cancellation
    Evaluation MethodologyConstruct ValidityBenchmarks+1
    · 13′
  20. C20
    Frontier AI Standards Body
    GovernanceAI PolicyFrontier Labs+2
    · 21′
  21. C21
    Expenditure Horizon
    BenchmarksCapability EvaluationEvaluation Methodology+3
    · 30′
  22. C22
    Agent-Authored Harness Optimization
    Agent EngineeringHarnessEvaluation+2
    · 127′
  23. C23
    Knowledge-Centric Self-Improvement
    Agent EngineeringKnowledge ManagementRecursive Self Improvement+2
    · 47′
  24. C24
    Orchestration-Plan Simulation
    BenchmarksEvaluation MethodologyMulti Agent+2
    · 25′
  25. C25
    Benchmark Contamination and Decontamination
    LLM ArchitectureCapability EvaluationBenchmarks+4
    · 30′
  26. C26
    Benchmark Score Redundancy
    LLM ArchitectureCapability EvaluationBenchmarks+3
    · 49′
  27. C27
    Measuring Beyond Accuracy Saturation
    LLM ArchitectureCapability EvaluationBenchmarks+3
    · 62′
  28. C28
    Reference-Free Judge Over-Crediting
    EvaluationLLM As A JudgeBenchmarks+3
    · 46′
  29. C29
    LLM-Judge Validation
    EvaluationLLM As A JudgeBenchmarks+2
    · 86′
  30. C30
    Compute-Controlled Benchmarking
    LLM ArchitectureCapability EvaluationBenchmarks+2
    · 82′
  31. C31
    DRACO Benchmark
    BenchmarksCapability EvaluationDeep Research+1
    · 13′
  32. C32
    LLM-as-a-Judge
    EvaluationLLM As A JudgeBenchmarks+1
    · 67′
  33. C33
    Production-Sourced Evaluation
    EvaluationBenchmarksData Provenance+1
    · 32′
  34. E34
    METR
    EntityOrgAI Evaluation+1
    · 25′
  35. C35
    Task Time-Horizon Scaling
    LLM ArchitectureCapability EvaluationBenchmarks+1
    · 28′