LetTokenBurnAgenticBBO-Bench: Benchmarking LLM Agents for Black-Box Optimization
A cross-domain benchmark evaluating LLM agents on five black-box optimization tasks (synthetic, HPO, DB tuning, chip design, molecular) under a unified finite-budget protocol.
$ npm i -g lettokenburn-cli$ ltb bench info agenticbbo-bench
External harness. No Harbor adapter has been generated for this benchmark yet. `ltb bench info` prints the known links.
Runs happen on your machine with your own subscription; only the score summary is pushed. CLI docs
| Harness | Source | Confidence | Link | |||
|---|---|---|---|---|---|---|
| GPT-6 Astra | 0.6 | N/A | Independent | medium | Oct 8, 2026 | arxiv.org |
| Gemini 3.1 Pro | 0.6 | N/A | Independent | medium | Oct 8, 2026 | arxiv.org |
| GPT-5.6 Sol | 0.6 | N/A | Independent | medium | Oct 8, 2026 | arxiv.org |
| Kimi K3 | 0.6 | N/A | Independent | medium | Oct 8, 2026 | arxiv.org |
| DeepSeek V4.1 Flash | 0.6 | N/A | Independent | medium | Oct 8, 2026 | arxiv.org |
| GLM-5.3 | 0.5 | N/A | Independent | medium | Oct 8, 2026 | arxiv.org |
| DeepSeek V4 Pro | 0.5 | N/A | Independent | medium | Oct 8, 2026 | arxiv.org |
Every number links to where it was published. Low-confidence rows are muted and need per-cell verification.
AgenticBBO-Bench covers synthetic functions, hyperparameter optimization, database tuning, chip design, and molecular design. It uses a unified finite-budget evaluation protocol to compare agentic systems against numerical optimizers.
Scoring relies on family-averaged scores across domains. Results show agentic BBO outperforms direct LLM methods in all domains and best numerical optimizers in four, with GPT-6 Astra and DeepSeek-V4.1-Flash on the Pareto frontier.