CodingDeprecated
Trust Score:
10

CodeContests

DeepMind's competitive-programming corpus with temporally split problems, human submissions, and generated correctness tests.

Launched: Refresh: static
Status Assessment (deprecated):

The repository was archived on 2024-12-06, the temporal holdout has been public since 2022, and later AlphaCode systems used updated non-public data; it no longer provides a clean maintained frontier comparison.

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 (72.0%)
ModelScoreDateSource TypeProvenance
AlphaCode 9B (test set, 10@100k, no clustering)25.8%2022-02-02vendor-reportedSource ↗
AlphaCode 41B (test set, 10@100k, no clustering)27.7%2022-02-02vendor-reportedSource ↗
AlphaCode 41B + clustering (test set, 10@100k)29.6%2022-02-02vendor-reportedSource ↗

Human Baseline & Difficulty Horizon

Calibrated human reference points, specialist benchmarks, and ceiling thresholds.
Measured Human Score72.0%Domain Expert Baseline
Baseline Protocol & Interpretation

Average solve rate among rated Codeforces and CodeChef human competitive programmers.

Metric & Scoring Methodology

Verification protocols, aggregation formulas, and specialized metric variants.
Primary Metric:10@100k solve rate on the 165-problem test split (%)
Scoring Engine:unit-tests

Dataset & Compute Cost

Evaluation volume, public availability, API pricing, and local hardware requirements.
Total Dataset Size165Annotated 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 Verifierunit-tests
Option AEleutherAI LM-Evaluation-Harness (Open-Weight Models)
lm_eval --model hf --model_args pretrained=<model_path> --tasks codecontests --batch_size auto
Option BOpenCompass Evaluation Framework
opencompass --datasets codecontests --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 CodeContests, 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