Research map / Learning theory and algorithms / Sample complexity
Limiting Extrapolation in Linear Approximate Value Iteration
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 LAVIER interpolation (min l1 interpolation weights over anchors) instead of least-squares AVI | LS-AVI · Two-state MDP (Tsitsiklis-Van Roy) | prediction error vs horizon | Helped |
| use LAVIER with greedy anchors (C<=1.2) instead of LS-AVI | LS-AVI · Chain MDP N=50 | suboptimality at s1 | Helped |
| LAVIER with small C (1.05, 1.2) instead of LS-AVI | LS-AVI · Successive linear bandits | V*(st) error | Helped |
| loosen amplification target from C=1.05 to C=1.5 (fewer anchors) (ablation) | LAVIER · Successive linear bandits | V*(st) error | Mixed |
| use LSTD off-policy prediction instead of LAVIER | LAVIER · Successive linear bandits | value estimate quality vs chain length | Hurt |
Closest papers by meaning
- Least squares variational inference (NeurIPS 2025 · similarity 0.84)
- Nearly Optimal Approximation of Matrix Functions by the Lanczos Method (NeurIPS 2024 · similarity 0.84)
- EigenVI: score-based variational inference with orthogonal function expansions (NeurIPS 2024 · similarity 0.84)
- Minimum Norm Interpolation Meets The Local Theory of Banach Spaces (ICML 2024 · similarity 0.83)
- Towards Learning High-Precision Least Squares Algorithms with Sequence Models (ICLR 2025 · similarity 0.83)
- Fast Last-Iterate Convergence of SGD in the Smooth Interpolation Regime (NeurIPS 2025 · similarity 0.83)
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.