MathematicsActive
Trust Score:
95

Soohak

Fresh research-level mathematics authored by mathematicians, plus unanswerable prompts that reward justified refusal.

Launched: Refresh: static
Status Assessment (active):

The official July 2026 tracker reports 64.4% Avg@3 for the Challenge leader with 46 of 340 problems still unsolved by any tracked model; Refusal remains a separate hard target.

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 (50.6%)
ModelScoreDateSource TypeProvenance
Gemini 3 Pro (Challenge Avg@3)30.39%2024-06-01independentSource ↗
GPT-5.5 high (Challenge Avg@3)64.4%2024-06-01independentSource ↗

Human Baseline & Difficulty Horizon

Calibrated human reference points, specialist benchmarks, and ceiling thresholds.
Measured Human Score50.6%Domain Expert Baseline
Baseline Protocol & Interpretation

Collaborative solve rate achieved by 5 invited competitive mathematics human teams.

Metric & Scoring Methodology

Verification protocols, aggregation formulas, and specialized metric variants.
Primary Metric:Challenge Avg@3 (%)
Scoring Engine:llm-judge

Dataset & Compute Cost

Evaluation volume, public availability, API pricing, and local hardware requirements.
Total Dataset Size439Annotated 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 Verifierllm-judge
Option AEleutherAI LM-Evaluation-Harness (Open-Weight Models)
lm_eval --model hf --model_args pretrained=<model_path> --tasks soohak --batch_size auto
Option BOpenCompass Evaluation Framework
opencompass --datasets soohak --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 Soohak, 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:LOW
Refresh Cadence:

Static fixed snapshot

Test Set Exposure:

Public on web / HuggingFace