Mathematics

Group Relative Policy Optimization (GRPO): Mathematical Foundations, Value-Free Advantage Estimation, Group Normalization Dynamics, and Scalable Reasoning RL

Post-training reinforcement learning (RL) has become the core driver of complex reasoning capabilities in frontier language models. While early alignment workflows focused on conversational preference modeling via Proximal Policy Optimization (PPO) or Direct Preference Optimization (DPO), scaling reinforcement learning to multi-step mathematical derivation and code generation revealed structural inefficiencies in classical Actor-Critic architectures. The primary operational constraint in tradit

The Muon Optimizer: Mathematical Foundations, Newton-Schulz Polar Iterations, and Spectral Steepest Descent in LLM Pre-Training

Transformer pre-training has long relied on AdamW as the default first-order optimizer. While AdamW provides coordinate-wise scale invariance through running estimates of first and second gradient moments, it treats every weight parameter as an isolated scalar. For modern neural networks composed primarily of two-dimensional linear projections, attention transformations, and feed-forward weight matrices, this coordinate-wise factorization neglects the structural transformation properties of matr

Decoupled Weight Decay Regularization (AdamW): Mathematical Foundations, L2 Regularization Failure in Adaptive Gradient Methods, and Hyperparameter Dynamics

Decoupled weight decay regularization, formalized by Ilya Loshchilov and Frank Hutter in 2017 and published at ICLR 2019, resolved a fundamental flaw in how adaptive gradient algorithms like Adam implemented weight regularization. Prior to AdamW, deep learning frameworks implemented weight decay as L2 regularization by adding the gradient of the squared L2 norm directly to the loss gradient. While mathematically identical for standard stochastic gradient descent (SGD), this formulation breaks do

Wasserstein GANs: How Earth Mover's Distance, Kantorovich-Rubinstein Duality, and Gradient Penalties Fixed Generative Adversarial Training

Wasserstein GANs: How Earth Mover's Distance, Kantorovich-Rubinstein Duality, and Gradient Penalties Fixed Generative Adversarial Training Generative Adversarial Networks (Goodfellow et al., 2014) established an influential paradigm for generative modeling: formulating data synthesis as a minimax game between a generator mapping latent noise to data space and a discriminator distinguishing synthetic samples from real empirical data. Despite early empirical success, the original GAN formulation

Neural Ordinary Differential Equations: How Continuous-Depth Dynamics and Adjoint Sensitivity Solve the Memory Bottleneck in Deep Learning

Neural Ordinary Differential Equations: How Continuous-Depth Dynamics and Adjoint Sensitivity Solve the Memory Bottleneck in Deep Learning Deep neural networks are traditionally structured as a discrete sequence of layers. An input tensor passes through layer after layer, transforming its representation at fixed, integer time steps. In standard architectures like Residual Networks (ResNets), each successive block computes an additive update: h_{t+1} = h_t + f(h_t, \theta_t) In 2018, researche

Score-Based Generative Modeling via Stochastic Differential Equations: How Continuous SDEs and Score Matching Unify Diffusion Models

Diffusion and score-based models represent one of the foundational paradigms of modern generative artificial intelligence, underpinning systems across image synthesis, video generation, audio modeling, and continuous multimodal representations. For years, generative diffusion was approached from two distinct perspectives: discrete-step denoising diffusion probabilistic models (DDPM) pioneered by Sohl-Dickstein et al. and Ho et al., and score matching with Langevin dynamics (SMLD / NCSN) introduc

Terence Tao Warns AI-Driven Proof Abundance Risks Mathematical Comprehension Crisis

In a paper prepared for the 2026 International Congress of Mathematicians, mathematician Terence Tao argues that artificial intelligence will force a restructuring of mathematical research practices, publication criteria, and education. The essay, released on arXiv (2608.16753), outlines how the transition from proof scarcity to proof abundance creates operational and epistemological challenges distinct from earlier debates over automated theorem proving. Tao frames the incoming disruption agai

An AI system produced a machine-checked proof of the 246 prime-gaps theorem

Neurosurgery resident solves Crouzeix's conjecture with GPT-5.6

A neurosurgery resident with no formal advanced mathematics training has solved Crouzeix's conjecture, a 22-year-old open problem in numerical linear algebra, using a 16-hour autonomous run of GPT-5.6 Sol in ChatGPT Work mode. Dr. Shanmu Jin, a postdoctoral researcher and neurosurgery resident at Peking Union Medical College Hospital, prompted ChatGPT 5.6 on July 30, 2026 to work on the conjecture. The model ran autonomously for approximately 16 hours, exploring proof strategies through a branc

Claude Advances Riemann Zeta Function Research, Raising Zero Point Bound to 67.2%

Claude Advances Riemann Zeta Function Research, Raising Zero Point Bound to 67.2%

OpenAI says Astra solved ten open math problems for about $2,000

OpenAI has named its next major model, Astra, by publishing ten mathematical results it says were generated by an internal version of the system. The company released a technical paper, reasoning walkthroughs and Lean certificates that allow the formal proofs to be checked by software. The more important part of the announcement is not the count. AI-generated mathematics has produced enough false starts that a headline number alone is weak evidence. OpenAI's decision to publish machine-checkabl