Self-Generated Feedback Destabilizes Test-Time Training: A Causal Decomposition of Long-Horizon Adaptation
実装難易度
Easy
推論・学習コスト
High
想定用途
生成
概要
Test-time training (TTT) lets a model store information in its weights during inference. When the model learns from its own output, however, each update also changes the model that generates the next training example. Across 128K-token streams, retaining generated-text updates worsens prediction on independent human-written text with three TTT-E2E model configurations (labeled 125M, 760M, and 3B).…
何が新しいか
Test-time training (TTT) lets a model store information in its weights during inference. When the model learns from its own output, however, each update also changes the model that generates the next…
何に使えるか
生成
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
SimForcing: Distilling Simulation Motion Priors into Real-Domain Robot World Models
Action-conditioned robot world models must respond precisely to robot trajectories while preserving realistic
LoGRA: Scaling LLM Reinforcement Learning with Low-Rank Gradient Sketches
Reinforcement learning (RL) has greatly advanced the capabilities of large language models (LLMs), but its mem
Multi-Behavioral Evolved Substrates Through Neuromodulation and Activation Selection
Open-ended artificial life systems must acquire diverse competencies from a single evolving genotype. Biologic
WASD: Wasserstein-based Knowledge Distillation for Large Language Models
Autoregressive large language models (LLMs) have rapidly advanced in capability, but their increasing scale co