Skip to content
BenchProof

Case file · worked example

FINALIZEDMMLUScenario C

Maverick posts 89.1 on MMLU

Llama-4-Maverick achieves 89.1 on MMLU.

Insufficient

The score is not backed by enough methodology or artifacts.

01 · What was claimed

The published sentence

Model A
Llama-4-Maverick · unknown
Model B / baseline
unspecified baseline · n/a
Benchmark
MMLU · unspecified
Evaluation date
2026-01-09
Metric
accuracy
Reported result
89.1
Claimant
0xC0FFEE0000000000000000000000000000C0FF
Published
2026-01-09T16:05:00.000Z

Methodology

Score posted on social media.

Retries

— vs —

Temperature

— vs —

Tools

— vs —

Prompts

undisclosed

System prompts

undisclosed

Sample

undisclosed

02 · What evidence was provided

Off-chain artifacts, on-chain hashes

Large files are not stored in the Intelligent Contract. The contract holds canonical hashes and references. The index below is the case file.

  • note

    Screenshot of a score. No split, no harness, no dataset revision, no baseline.

    sha256:6e992beafafae82f9aec8282d3a95129d20116e274225195e0b8034b7539c12f

03 · What was challenged

Public review

  1. Challenge 01 · Reproducibility problem

    Nothing to reproduce

    No dataset version, no 5-shot versus 0-shot disclosure, no evaluation harness, no raw item scores. A number is not a benchmark result.

    0xAUD170000000000000000000000000000AUD17

04 · What was decided

Verdict

Insufficienthigh confidenceWorked example (same rubric)

The published score for Llama-4-Maverick vs unspecified baseline on MMLU cannot be verified from the submitted evidence and methodology.

Key findings

  • Compared the published claim against attached evidence and condition fields.

Material issues

  • Methodology is missing or too thin to reproduce the reported result.
  • Raw results or harness artifacts were not provided.

Limitations

  • Prompt and system-prompt parity were not fully disclosed.
  • This preview uses the BenchProof forensic rubric. The canonical verdict is the GenLayer consensus result.

05 · Verification

On-chain identifiers

Claim id
seed-insufficient
On-chain id
not submitted
Network
studionet · chain 61999
Tx state
finalized
Transaction
—

Off-chain rows never override a finalized on-chain verdict. Seeded case files illustrate the rubric; live GenLayer consensus is marked explicitly. Open contract in Studio · Live evaluation tx