MiniMax
24stories this week
28last 30 days
37all time
Timeline
- Oct 8, 2026 · Research paper · 2 sourcesWorldGuide: Goal-Directed Video World Model for Procedural Task ExecutionWe formulate procedural video generation as closed-loop task execution in visual world space and introduce WorldGuide.
- Oct 8, 2026 · Research paper · 1 sourceSubspace Uncertainty and Sharp Sampling Thresholds on the Boolean CubeWe study Gaussian regression under squared population $L2$ loss in a known $m$-dimensional subspace of degree-at-most-$k$ functions on the $d$-dimensional Boolean cube.
- Oct 8, 2026 · Research paper · 1 sourceTowards Unified Evaluation of Prompt Enhancers for Video GenerationTo address this gap, we introduce PEBench, the first unified benchmark for direct PE evaluation across text-to-video, image-to-video, and reference-to-video prompt enhancement.
- Oct 8, 2026 · Research paper · 1 sourceMinimax Gaussian Mechanisms for Continual Machine UnlearningWe develop Gaussian mechanisms for Newton updates under sequential deletion requests.
- Oct 8, 2026 · Research paper · 1 sourceNew Lower Bound and Upper Bounds on the Regret for Online Sparse Linear RegressionWe study online sparse linear regression (OSLR) where any algorithm is restricted to accessing only $b$ out of $d$ attributes per instance for prediction and $b0\geq 0$ additional attributes after prediction, which was proved to be NP-hard.
- Oct 8, 2026 · Research paper · 1 sourceParametric Trajectory Distillation for Few-Step Video GenerationWe introduce Parametric Trajectory Distillation (PTD), which lets the student parameterize teacher trajectory segments as polynomials and learn from teacher guidance along its own predicted path.
- Oct 8, 2026 · Research paper · 1 sourceBRACE: Differential Privacy for Dense Associative Memory with LSR EnergyIn this paper, we develop a differential privacy framework for log-sum-ReLU (LSR) dense associative memory, whose finite-support retrieval dynamics pose distinctive challenges for privacy-preserving computation.
- Oct 8, 2026 · Research paper · 1 sourceA General $\widetildeΩ(\sqrt{T γ_T})$ Lower Bound for Kernel BanditsThe kernel bandit problem consists of sequentially optimizing an unknown function with noisy feedback, where the function has bounded norm in a given Reproducing Kernel Hilbert Space (RKHS).
- Oct 7, 2026 · Research paper · 1 sourceData Reuse in Non-Stationary LearningWe consider online learning in non-stationary environments, where the goal is to track an unknown parameter that switches abruptly between a finite set of recurring values.
- Oct 7, 2026 · Research paper · 1 sourcePolicy Learning with Weak SignalsPolicy learning in digital experimentation faces three challenges: weak signal-to-noise ratios, rich covariate spaces, and massive data volumes.
- Oct 7, 2026 · Research paper · 1 sourceSharp Asymptotic Theory of Maximum Likelihood Estimation for Gaussian Processes with an RBF KernelGaussian processes (GPs) are widely used across machine learning, spatial statistics, time-series analysis, optimization, Bayesian statistics, and scientific applications.
- Oct 7, 2026 · Research paper · 1 sourceOnline Resource Allocation with an Endogenous Markov State: Fewer LP Solves Earn MoreWe study finite-horizon online resource allocation with i.i.d. requests and an endogenous Markov state on a finite state space: each action affects the transition of the state that governs future rewards and resource consumption.
- Oct 7, 2026 · Research paper · 1 sourceFinite-Rank Logistic Gaussian Processes with Exact Likelihood for Conditional Density EstimationWe propose the exact likelihood finite-rank LGP (ExFR-LGP), which writes the log density as the sum of two bivariate functions, one of the response and a location-varying linear index of the covariates, and one of the response and the location.
- Oct 6, 2026 · Research paper · 1 sourceSketched Calibration for Conformal Prediction under Covariate ShiftWeighted conformal prediction corrects for covariate shift by reweighting calibration scores with the likelihood ratio between target and source covariates.
- Oct 6, 2026 · Research paper · 1 sourceHoldout Best-of-N: Unbiased Evaluation and Its CostWe study evaluation from a fixed matrix of $K$ independent scores per candidate for a policy that selects using $J$ fresh scores.
- Oct 6, 2026 · Research paper · 1 sourceReinforcement Learning for Hierarchical Reasoning Rewards: Minimax-Optimal Rates with TransformersReinforcement learning (RL) has become a standard tool for post-training language models on reasoning tasks, where the policy is updated by reward feedback while exploring the space of responses.
- Oct 6, 2026 · Research paper · 1 sourceTwo-Sample Testing via Generative ProcessesDeciding whether two samples come from the same distribution is a classical problem in statistics, and generative transport offers a new way to approach it.
- Oct 6, 2026 · Research paper · 1 sourceDetecting a Shift Is Not Enough: Exact Minimax Limits of Linear Representation RepairWe cast its removal as a statistical decision problem: from noisy differences between paired calibration measurements in $\mathbb{R}^d$, learn one linear map, applied to both sources under a hard distortion budget, that leaves as little of the shift as possible on fresh data.
- Oct 6, 2026 · Open-source release · 1 sourcehuggingface/diffusers v0.41.0: Diffusers 0.41.0: QwenImage 2.1 pipeline and more> This release brings Qwen-Image 2.1 to Diffusers, with text-to-image generation, image editing, native transparency, and LoRA training.
- Oct 6, 2026 · Research paper · 1 sourceAdaptive Mean Estimation by In-Context Learning: A Gradient-Flow AnalysisPrior Fitted Networks (PFNs) such as TabPFN now rival established statistical procedures across prediction and estimation tasks.
- Oct 6, 2026 · Research paper · 1 sourceSlow Beats Fast at the Kesten-Stigum Threshold: Minimax, Fisher-Information and Belief-Propagation Characterizations of the Information-Computation Gap in Sparse Stochastic Block ModelsWe study community recovery in the sparse symmetric stochastic block model with $q$ communities, average degree $d$ and signal strength $λ$ through statistical decision theory and Fisher information, and obtain three characterizations of the Kesten-Stigum threshold $dλ^2=1$ and of the information-computation gap below it.
- Oct 6, 2026 · Research paper · 1 sourceNash Social Welfare for Multi Armed Bandits: Trajectory-wise Expected and High Probability RegretWe study fair multi-armed bandits under the Nash Social Welfare (NSW) objective, which measures performance via the geometric mean of accumulated rewards.
- Oct 6, 2026 · Research paper · 1 sourceUniform Discrete Diffusion Models are Minimax Optimal for Estimating Distributions with Small Effective Support SizeDiscrete diffusion models have emerged as a practically successful framework for generative modeling on discrete product spaces, yet their statistical generalization properties remain poorly understood.
- Oct 5, 2026 · Research paper · 1 sourceMC-Sparse: Deconstructing and Closing the Dense-Sparse Attention Gap in Diffusion TransformersSparse attention is a primary approach to reducing the latency of diffusion transformers in long-sequence generation tasks, such as video and high-resolution 3D asset generation.
- Oct 2, 2026 · Open-source release · 1 sourcesgl-project/sglang v0.5.21| Model | Type | Cookbook |
- Sep 29, 2026 · Open-source release · 1 sourceNVIDIA/TensorRT-LLM v1.3.0rc29Expose Nemotron-H vision-language LoRA configuration for supported inference paths #19151
- Sep 28, 2026 · Open-source release · 1 sourceunslothai/unsloth v0.1.900-beta: Laya Decision Models + LibraryRun and serve Decision Models like Laya (open-source Jev) locally
- Sep 18, 2026 · Open-source release · 1 sourcesgl-project/sglang v0.5.20| Model | Type | PRs | Cookbook |
- Sep 5, 2026 · Open-source release · 1 sourcesgl-project/sglang v0.5.19| Model | Type | PRs | Cookbook |
- Aug 10, 2026 · Open-source release · 1 sourcevllm-project/vllm v0.27.0Kimi K3 support with a full stack landing in one release: core model files and kernels (#50089, #50000), Python (#50093) and Rust (#50104) frontends, AttnRes kernels (#50090), DeepGEMM support (#50458), compressed-tensors quantized checkpoints (#50500), DSpark AR fusion (#50242), and an option to shard the shared expert instead of replicating it (#50656).