Can Terminal Agents Trust Their Own Verification? Diagnosing and Improving Self-Verification
Yingfeng Luo, Shaowei Wei, Daixin Wang, Dingyang Lin, Kaiyan Chang
実装難易度
Hard
推論・学習コスト
Medium
想定用途
異常検知
概要
Terminal agents rely on self-verification to assess and correct their solutions as they solve tasks through interaction with command-line environments. Yet how trustworthy such self-verification is remains poorly understood. To investigate this question systematically, we introduce a diagnostic framework that identifies the first complete solution in each trajectory, determines whether it is…
何が新しいか
Terminal agents rely on self-verification to assess and correct their solutions as they solve tasks through interaction with command-line environments. Yet how trustworthy such self-verification is…
何に使えるか
異常検知
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
未取得推論中心なら軽めですが、再学習時はGPUが必要になる可能性があります。
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: Medium
- 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。
- 推論中心なら軽めですが、再学習時はGPUが必要になる可能性があります。
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
Training LLM Judges from Language Feedback via Position-Selective Self-Distillation
We study training LLM judges from natural language feedback, especially for subjective tasks where the verdict
SparseEngine: Sparse-First Inference Engine
Long-context LLM agents accumulate interaction histories that strain KV-cache memory and attention computation
Recovering Off-Policy Supervision for Speculative Decoding
Block drafters for speculative decoding are commonly trained on corpora written by external models, where a si
Reinforcing Multimodal Reasoning via Token-Level Perception-Grounded Advantage Estimation
Reinforcement Learning with Verifiable Rewards (RLVR) has improved the reasoning capabilities of Multimodal La