NeurIPS 2026 reviews are up: reviews, scores, rebuttals and meta-reviews for 7,877 papers. See the highest rated
acceptodds: the papers of ICLR 2027
148 results for “keyword:"matrix"” in ICLR 2027. 24 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]ChromaMuon: Spectral Optimizer Diagnostics and Design with Retained Newton–Schulz Coordinates▼1.2k1.67
[PDF]Optimal Explicit and Implicit Regularization Strengths in High-Dimensional Continual Linear Regression▲5.4k3.00
[PDF]Decoupled Polynomial State-Space Iterations for Matrix Polar Approximation, with an Application to Muon▼4.3k2.00
[PDF]Sketch-Aware Generalized Linear Bandits: Efficient One-Pass Updates via Frequent Directions▼4.5k2.67
[PDF]Non-Asymptotic Last-Iterate Convergence for Scale-Invariant Regret-Matching Algorithms in Matrix Games▼5.8k2.33
[PDF]In-context classification: Calibration and noise in linear attention with softmax readout▼4.5k2.67
[PDF]Root Multiplicity Sets the Finite-Width Edge of Chaos in MLPs and Message-Passing Networks▼5.8k2.33
[PDF]Reconstruction Does Not Determine Generative Learnability: A Theory of Resource-Matched Visual Tokenization▼4.3k2.00
[PDF]Compression Hacking: Towards a Unified Informatics and Geometric Perspective on Intrinsic Evaluation of Large Language Models▲5103.00
[PDF]Smooth Spectral Operators for Indefinite Hessians: Relative Resolution, Stability, and Matrix-Free Approximation▼1611.50
[PDF]Spectral Dynamics in Deep Networks: Feature Learning, Outlier Escape, and Learning Rate Transfer▼1.2k3.00
Withdrawn or desk rejected (24)
[PDF]Training-Loss Guarantees for Muon with Finite-Step Newton-Schulz OrthogonalizationWithdrawn1.67
[PDF]Learning Scalar Quadratic Contrasts for High-Dimensional Change Point LocalizationDesk rejected1.75