What LLM Trading Agents Actually Do in Production: A Six-Month, Population-Scale Record from Two Fleets
実装難易度
Easy
推論・学習コスト
High
想定用途
回帰
概要
We present a continuous, population-scale measurement record of autonomous language-model trading agents operating in production across two systems with one design lineage: DX Terminal Pro (3,505 user-funded vaults trading real ETH in Base memecoin markets for 21 days, February to March 2026) and the DXAP live alpha fleet (500 to 599 user-created agents all-history, 91 to 117 concurrently active,…
何が新しいか
We present a continuous, population-scale measurement record of autonomous language-model trading agents operating in production across two systems with one design lineage: DX Terminal Pro (3,505…
何に使えるか
回帰
実装情報
- Hugging Face URL
- あり
実装チェックリスト
実装または配布ページ
OKコードまたはモデル配布ページから検証を始められます。
一次情報リンク
OKHugging Face
検証しやすさ
OK実装またはモデル配布ページから試せる可能性が高いです。
計算資源
要確認学習や高解像度推論ではGPUメモリと実行時間に注意が必要です。
ライセンス
未取得配布元のLICENSE、モデルカード、Paperの利用条件を確認してください。
商用利用
未取得研究利用限定、データセット由来制限、API規約の有無を確認してください。
自社データで試すなら
製造業・材料開発のExcel/CSVデータに落とし込むための最初の手順です。
- 1まず自社データを、入力条件、目的変数、評価したい指標に分けて整理します。
- 2LightGBMやRandom Forestなどのベースラインを先に作り、この手法と比較します。
- 3評価指標はR2/RMSE、AUC、異常検知の再現率、実験回数削減率など、現場の意思決定に近いものを選びます。
- 4SHAPや特徴量重要度で、効いている因子が物理・化学・工程知識と矛盾しないか確認します。
実装難易度
Easy - 実装またはモデル配布ページから試せる可能性が高いです。
必要リソース
- GPU目安: High
- データセット: 論文・リポジトリ側の指定を確認してください。
- 学習要否: 推論だけで試せる可能性があります。
- 学習や高解像度推論ではGPUメモリと実行時間に注意が必要です。
実務で使う場合の注意点
- ライセンスと商用利用条件は、Paper / GitHub / Hugging Face の配布元で確認してください。
- 精度、再現性、計算コストはデータセットや評価条件に依存します。
- 個人情報や機密データを扱う場合は、入力データの保存先と外部API利用条件を確認してください。
関連記事
smile — Statistical Machine Intelligence & Learning Engine
マシン学習、統計学習などに関する統計的エンジンです。
CausalVerify: An Execution-Grounded Benchmark for LLM Causal Inference Workflows
Existing causal-inference benchmarks for LLMs mostly score method descriptions or whether generated code runs,
Automating Parent Selection Configuration in Genetic Programming with Agentic AI
We investigate whether agentic artificial intelligence can automate parts of the process of designing genetic
Benchmarking Zero-Shot LLM-Generated Parent Selection in Genetic Programming for Symbolic Regression
Parent selection significantly affects exploration, exploitation, and complexity control in genetic programmin