ARCBench S3 - LRU (rule) Full Context: leaderboard
ARCBench pi agent harness; same exact endpoint is agent and compactor, 17 task cases with only configurations having all valid repetitions included. Violation@k is percent of cases with at least one violation in k runs; lower is better. GPT-Luna retains the recorded mutable ~openai/gpt-luna-latest endpoint without an inferred dated identity. S3; k=3, channels c1/c3; LRU (rule); Full Context; no attack or defense. Valid-run subsets may vary by strategy.
Metric: Violation@3 (%). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | gpt-luna-latest (ARCBench Run: endpoint=~openai%2Fgpt-luna-latest; harness=pi; strategy=LRU%20%28rule%29; context=Full%20Context) | 5.9 | |
| 2 | deepseek-v4-flash-0731 (ARCBench Run: endpoint=deepseek%2Fdeepseek-v4-flash-0731; harness=pi; strategy=LRU%20%28rule%29; context=Full%20Context) | 11.8 | |
| 3 | glm-5.3-flash (ARCBench Run: endpoint=z-ai%2Fglm-5.3-flash; harness=pi; strategy=LRU%20%28rule%29; context=Full%20Context) | 11.8 | |
| 4 | qwen3.7-flash (ARCBench Run: endpoint=qwen%2Fqwen3.7-flash; harness=pi; strategy=LRU%20%28rule%29; context=Full%20Context) | 17.6 |
Interactive version: theaggregate.ai/benchmark?slug=arcbench-s3-lru-rule-full-context · How It Works · Data refreshed daily, snapshot 2026-11-05.