SWE-bench
Real GitHub issues from 12 Python repositories, solved by generating patches that must pass repository tests.
SWE-bench Verified superseded the original test set in 2024 after human review found underspecified tasks and tests that rejected valid fixes.
Performance Timeline
Longitudinal progression of model scores against human baselines.Performance & Historical Trajectory
Empirical score progression across model release dates and evaluation rounds.
Human Baseline & Difficulty Horizon
Calibrated human reference points, specialist benchmarks, and ceiling thresholds.Measured solve rate of experienced software engineers resolving unfamiliar repository GitHub issues within standard 4-hour time limits.
Metric & Scoring Methodology
Verification protocols, aggregation formulas, and specialized metric variants.resolved instances (%)Dataset & Compute Cost
Evaluation volume, public availability, API pricing, and local hardware requirements.$5 – $20 USD for full benchmark evaluation run on frontier APIs.
1x NVIDIA RTX 4090 (24GB) or A100 (40GB/80GB) via vLLM / SGLang
How to Run & Reproduce
Standardized evaluation protocols, CLI commands, and reproducible runner templates.lm_eval --model hf --model_args pretrained=<model_path> --tasks swe-bench --batch_size autoopencompass --datasets swe-bench --models <model_config># 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,
)When publishing results for SWE-bench, 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.Static fixed snapshot
Public on web / HuggingFace