ReasoningSaturated
Trust Score:
25

SuperGLUE

Eight English NLU tasks spanning question answering, inference, causal reasoning, word sense, and coreference.

Launched: Refresh: static
Status Assessment (saturated):

The official aggregate human estimate (89.8) was crossed by a single DeBERTa model (89.9) by 2020-12-29; the 2022 leaderboard leader reached 91.3, leaving a narrow, noisy ceiling.

Performance Timeline

Longitudinal progression of model scores against human baselines.

Performance & Historical Trajectory

Empirical score progression across model release dates and evaluation rounds.

Independent Vendor-Reported Human Baseline (89.8%)
ModelScoreDateSource TypeProvenance
Vega v291.3%2022-10-08vendor-reportedSource ↗
PaLM 540B90.4%2022-04-04vendor-reportedSource ↗
ST-MoE-32B91.2%2022-02-17vendor-reportedSource ↗
ERNIE 3.090.6%2021-07-12vendor-reportedSource ↗
DeBERTa ensemble (TuringNLRv4)90.3%2021-01-06vendor-reportedSource ↗
DeBERTa 1.5B (single model)89.9%2020-12-29vendor-reportedSource ↗
T5-11B89.3%2019-10-23vendor-reportedSource ↗
BERT++71.5%2019-05-02independentSource ↗

Human Baseline & Difficulty Horizon

Calibrated human reference points, specialist benchmarks, and ceiling thresholds.
Measured Human Score89.8%Domain Expert Baseline
Baseline Protocol & Interpretation

Human crowdworker accuracy on SuperGLUE linguistic reasoning and entailment suite.

Metric & Scoring Methodology

Verification protocols, aggregation formulas, and specialized metric variants.
Primary Metric:SuperGLUE aggregate score (%)
Scoring Engine:composite

Dataset & Compute Cost

Evaluation volume, public availability, API pricing, and local hardware requirements.
Total Dataset SizeAnnotated evaluation items
Public Test SetPublicOpenly mirrored on repositories
Access GatingOpen AccessUnrestricted download
Evaluation LicenseOpen AccessDataset usage and redistribution terms
Frontier API Compute Cost:

$5 – $20 USD for full benchmark evaluation run on frontier APIs.

Recommended Local GPU Setup:

1x NVIDIA RTX 4090 (24GB) or A100 (40GB/80GB) via vLLM / SGLang

Official Dataset & Benchmark Files:Download / View Dataset Repository ↗

How to Run & Reproduce

Standardized evaluation protocols, CLI commands, and reproducible runner templates.
Prompt Regimezero-shot
Reasoning Modedirect
Sampling Temp0
Pass@k Budgetk = 1
Tools & SandboxPure Text
Scoring Verifiercomposite
Option AEleutherAI LM-Evaluation-Harness (Open-Weight Models)
lm_eval --model hf --model_args pretrained=<model_path> --tasks superglue --batch_size auto
Option BOpenCompass Evaluation Framework
opencompass --datasets superglue --models <model_config>
Python APIDeterministic Inference Loop Snippet
# Standard API Evaluation Loop
from openai import OpenAI

client = OpenAI()
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": prompt}],
    temperature=0.0,
)
Standardized Reporting Requirement:

When publishing results for SuperGLUE, always report the exact prompt template, few-shot exemplar ordering, sampling temperature (temperature=0), maximum reasoning budget tokens, and the precise timestamped model snapshot ID.

Contamination & Memorization Analysis

Audit of pretraining exposure risks, memorization vectors, and refresh policies.
Overall Contamination Risk:HIGH
Refresh Cadence:

Static fixed snapshot

Test Set Exposure:

Public on web / HuggingFace