Hindcast: Replaying Prediction Markets to Evaluate LLM Forecasters
Xiao Ye, Jacob Dineen, Evan Zhu, Shijie Lu, Kevin Song
実装難易度
Hard
推論・学習コスト
High
想定用途
予測
概要
Forecasters are evaluated by backtesting, which replays resolved questions and grades the probability the system would have assigned before the outcome was known. For LLMs, two channels leak the answer into this test. A model that retrieves can surface reports written after the event, turning forecasting into a lookup, and each new model is trained on data closer to the event, so a question that…
何が新しいか
Forecasters are evaluated by backtesting, which replays resolved questions and grades the probability the system would have assigned before the outcome was known. For LLMs, two channels leak the…
何に使えるか
予測
実装情報
- Paper URL
- あり
実装チェックリスト
実装または配布ページ
要確認Paper onlyの可能性があるため再実装前提で確認してください。
一次情報リンク
OKPaper
検証しやすさ
要確認公式実装が見つからないため、論文から再実装する前提です。
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
Hard - 公式実装が見つからないため、論文から再実装する前提です。
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: 再学習や評価環境の準備が必要にな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
What LLM Forecasters Know but Don't Say: Probing Internal Representations for Calibration and Faithfulness
Large language models fine-tuned for forecasting can be accurate yet poorly calibrated, and their chain-of-tho
Decentralized Aggregation of LLM Predictions via Wagering Mechanisms
It is increasingly common to aggregate predictions from multiple LLMs, each with domain expertise or access to
GFlowRL: Scaling Distribution-Matching RL to Large Language Models
難解決問題に対する推論モデルの学習を改善するための分布一致型の強化学習を開発します。この強化学習は、推論モデルの分布の一致を強化することで、新しい学習タスクに対して適応する能力が向上します。
Towards Enhancing 3D Spatial Reasoning in Medical Multimodal Large Language Models
3D 医学的多モーダル言語モデルの推論は、医学の多モーダル言語モデルの推論を含む医学的推論に不可欠です しかし、医学的推論は複雑なタスクであり、医学的多モーダル言語モデルの推論には複雑さが存在することを示しています