Research map / Learning theory and algorithms / Kernels and regression theory
A Decision-Theoretic View of Test-Time Training: When, How Far, and Which Directions to Adapt
Tomoya Wakayama
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use fixed-T TTT (T=8) instead of no-update ICL (ablation) | distilgpt2 TTT · Digit-shift task | query MSE: 0.0747 -> 0.0734 | Improved |
| select T per prompt by evidence (MLE-σ) instead of fixed T=8 (ablation) | distilgpt2 TTT · Digit-shift task | query MSE: 0.0734 -> 0.0719 | Improved |
| select T per prompt by evidence (fixed-σ plug-in) instead of fixed T=8 (ablation) | distilgpt2 TTT · Digit-shift task | query MSE: 0.0734 -> 0.0722 | Improved |
| estimate σ by MLE instead of fixed-σ plug-in in evidence selection (ablation) | evidence-based T selection · Digit-shift task | query MSE: 0.0722 -> 0.0719 | No clear effect |
| use evidence-selected T in noisy vs clean prompts | evidence-based T selection · Digit-shift task (noisy vs clean) | selected T | Improved |
| select heads with QUERY-AWARE score instead of random selection (ablation) | head/block selection TTT · Digit-shift task, fixed block budget | query MSE improvement vs random | Improved |
| select blocks by TRACE-TOPK (prompt-fit) instead of QUERY-AWARE (ablation) | head/block selection TTT · Digit-shift task, fixed block budget | query MSE | Hurt |
Closest papers by meaning
- A Layer Selection Approach to Test Time Adaptation (AAAI 2025 · similarity 0.89)
- Test-Time Training Provably Improves Transformers as In-context Learners (ICML 2025 · similarity 0.88)
- Backpropagation-Free Test-Time Adaptation via Probabilistic Gaussian Alignment (NeurIPS 2025 · similarity 0.88)
- Rethinking Expressivity and Efficiency in Test-Time Training (NeurIPS 2026 · similarity 0.88)
- Compute Aligned Training: Optimizing for Test Time Inference (NeurIPS 2026 · similarity 0.87)
- Evaluation of Test-Time Adaptation Under Computational Time Constraints (ICML 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
