arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Author

MiniMax

7 papers · 1995–2025

1995: 11996: 11997: 01998: 01999: 22000: 02001: 02002: 02003: 02004: 02005: 02006: 02007: 02008: 02009: 02010: 02011: 02012: 12013: 02014: 02015: 02016: 02017: 02018: 02019: 02020: 02021: 02022: 02023: 02024: 02025: 2
1995peak 2 in 19992025

Frequent co-authors

Papers

showing 1–7 of 7 RSS

MiniMax-01: Scaling Foundation Models with Lightning Attention

MiniMax, Aonian Li, Bangwei Gong +87 more

cs.CL · · cited 1× · arXiv ↗ · more like this ·

We introduce MiniMax-01 series, including MiniMax-Text-01 and MiniMax-VL-01, which are comparable to top-tier models while offering superior capabilities in processing longer contexts. The core lies in lightning attention and its efficient scaling. To maximize computational capacity, we integrate it with Mixture of Experts (MoE), creating a model with 32 experts and 456 billion total parameters, of which 45.9 billion are activated for each token. We develop an optimized parallel strategy and highly efficient computation-communication overlap techniques for MoE and lightning attention. This approach enables us to conduct efficient training and inference on models with hundreds of billions of parameters across contexts spanning millions of tokens. The context window of MiniMax-Text-01 can reach up to 1 million tokens during training and extrapolate to 4 million tokens during inference at…

MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention

MiniMax, :, Aili Chen +125 more

cs.CL · · cited 0× · arXiv ↗ · more like this ·

We introduce MiniMax-M1, the world's first open-weight, large-scale hybrid-attention reasoning model. MiniMax-M1 is powered by a hybrid Mixture-of-Experts (MoE) architecture combined with a lightning attention mechanism. The model is developed based on our previous MiniMax-Text-01 model, which contains a total of 456 billion parameters with 45.9 billion parameters activated per token. The M1 model natively supports a context length of 1 million tokens, 8x the context size of DeepSeek R1. Furthermore, the lightning attention mechanism in MiniMax-M1 enables efficient scaling of test-time compute. These properties make M1 particularly suitable for complex tasks that require processing long inputs and thinking extensively. MiniMax-M1 is trained using large-scale reinforcement learning (RL) on diverse problems including sandbox-based, real-world software engineering environments. In addition…

Approximate Confidence Regions for Minimax-Linear Estimators

Minimax-linear Estimators, H. Toutenburg, A. Fieger +1 more

Statistics and Probability · · cited 0× · OpenAlex ↗ · more like this ·

Minimax estimation is based on the idea, that the quadratic risk function for the estimate fi is not minimized over the entire parameter space IR K , but only over an area B(fi) that is restricted by a priori knowledge. If all restrictions define a convex area, this area can often be enclosed in an ellipsoid of the form B(fi) = ffi : fi 0 T fi rg. The ellipsoid has a larger volume than the cuboid. Hence, the transition to an ellipsoid as a priori information represents a weakening, but comes with an easier mathematical handling. Deriving the linear Minimax estimator we see that it is biased and nonoperationable. Using an approximation of the non-central 2 -distribution and prior information on the variance, we get an operationable solution which is compared with OLSE with respect to the size of the corresponding confidence intervals. 1 Introduction We consider the linear regression…

TOPOLOGICALLY FINITE INTERSECTION PROPERTY

Minimax Theorems, Zhang Shi-sheng, Xian‐Ming Zhang

Numerical Analysis · · cited 0× · OpenAlex ↗ · more like this ·

A more general topologically finite intersection property is obtained. As an application; we utili:e this result to obtain some more general minimax theorems. The results presented in this paper unify and extend the main results of (5, 6, 9). Abstract The one-dimensional problem of the motion of a rigid flying plate under explosive attack has an analytic solution only when the polytropic index of detonation products equals to three. In general, a numerical analysis is required. In this paper, however, by utilizing the weak shock behavior of the reflection shock in the explosive products, and applying the small parameter pur- terbation method, an analytic, first-order approximate solution is obtained for the problem of flying plate driven by various high explosives with polytropic indices other than but nearly equal to three. Final velocities of flying plate obtained agree very well with…

A Search for Disoriented Chiral Condensate at the Fermilab Tevatron

MiniMax Collaboration, T. C. Brooks, M. E. Convery +21 more

hep-ex · · arXiv ↗ · more like this ·

We present results from MiniMax (Fermilab T-864), a small test/experiment at the Tevatron designed to search for the production of disoriented chiral condensate (DCC) in $p - \bar p$ collisions at $\sqrt{s} = 1.8$ TeV in the forward direction, $\sim 3.4 < η< \sim 4.2$. Data, consisting of $1.3 \times 10^6$ events, are analyzed using the robust observables developed in an earlier paper. The results are consistent with generic, binomial-distribution partition of pions into charged and neutral species. Limits on DCC production in various models are presented.

Analysis of Charged-Particle/Photon Observables in Hadronic Multiparticle Production

MiniMax Collaboration

hep-ph · · arXiv ↗ · more like this ·

In order to analyze data on joint charged-particle/photon distributions from an experimental search (T-864, MiniMax) for disoriented chiral condensate (DCC) at the Fermilab Tevatron collider, we have identified robust observables, ratios of normalized bivariate factorial moments, with many desirable properties. These include insensitivity to many efficiency corrections and the details of the modeling of the primary pion production, and sensitivity to the production of DCC, as opposed to the generic, binomial-distribution partition of pions into charged and neutral species. The relevant formalism is developed and tested in Monte-Carlo simulations of the MiniMax experimental conditions.