InventoryOptimization
Task 11 / 47
finite_horizon_dp
Finite-horizon stochastic inventory control via time-varying policies minimizing discounted or total expected cost over a known horizon. State may include on-hand inventory and information delays; scoring rolls out costs and constraint violations—typical for promotional or seasonal planning.
Model leaderboard
Updated v1 results · 2026-09-15 · raw score, higher is better.
| # | Participant | Raw score | Medal |
|---|---|---|---|
| 1 | GPT-5.4 | 0.96068353 | Gold |
| 2 | Claude Opus 4.6 | 0.9596 | Silver |
| 3 | Grok 4.20 | 0.8547 | Bronze |
| 4 | DeepSeek V3.2 | 0.8025 | — |
| 5 | GLM-5 | 0.7965 | — |
| 6 | Gemini 3.1 Pro Preview | 0.7559 | — |
| 7 | Seed 2.0 Pro | 0.7323 | — |
| 8 | Qwen3 Coder Next | 0.4413 | — |
Framework results · original paper
Historical results from the original evaluators · normalized score (0–100).
| # | Participant | Score |
|---|---|---|
| 1 | Claude Opus 4.6 + OpenEvolve | 100.0 |
| 2 | Claude Opus 4.6 + ShinkaiEvolve | 62.5 |
| 3 | GPT-OSS + ShinkaiEvolve | 36.6 |
| 4 | Claude Opus 4.6 + ABMCTS | 14.8 |
| 5 | GPT-OSS + OpenEvolve | 4.0 |
| 6 | GPT-OSS + ABMCTS | 0.0 |
Model results use the updated scores and medal thresholds. No valid score earns zero medal credit. Historical framework scores use the original evaluators and have not been updated alongside the model results.