惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

B
Blog
D
Docker
J
Java Code Geeks
腾
腾讯CDC
Blog — PlanetScale
Blog — PlanetScale
G
Google Developers Blog
M
MIT News - Artificial intelligence
L
LangChain Blog
T
The Blog of Author Tim Ferriss
P
Proofpoint News Feed
MyScale Blog
MyScale Blog
博
博客园 - Franky
GbyAI
GbyAI
Hugging Face - Blog
Hugging Face - Blog
aimingoo的专栏
aimingoo的专栏
Last Week in AI
Last Week in AI
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
博
博客园 - 聂微东
N
Netflix TechBlog - Medium
B
Blog RSS Feed
Y
Y Combinator Blog
阮一峰的网络日志
阮一峰的网络日志
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
Google DeepMind News
Google DeepMind News

cs.LG updates on arXiv.org

Memory-Guided Trust-Region Bayesian Optimization (MG-TuRBO) for High Dimensions EngageTriBoost: Predictive Modeling of User Engagement in Digital Mental Health Intervention Using Explainable Machine Learning Reservoir observer enhanced with residual calibration and attention mechanism Efficient RL Training for LLMs with Experience Replay Wireless Communication Enhanced Value Decomposition for Multi-Agent Reinforcement Learning Adversarial Sensor Errors for Safe and Robust Wind Turbine Fleet Control IKKA: Inversion Classification via Critical Anomalies for Robust Visual Servoing Adaptive Simulation Experiment for LLM Policy Optimization EvoLen: Evolution-Guided Tokenization for DNA Language Model Smartwatch-Based Sitting Time Estimation in Real-World Office Settings Structural Evaluation Metrics for SVG Generation via Leave-One-Out Analysis Loom: A Scalable Analytical Neural Computer Architecture Spectral Geometry of LoRA Adapters Encodes Training Objective and Predicts Harmful Compliance Finite-Sample Analysis of Nonlinear Independent Component Analysis:Sample Complexity and Identifiability Bounds How does Chain of Thought decompose complex tasks? Uncertainty-Aware Transformers: Conformal Prediction for Language Models Adaptive Candidate Point Thompson Sampling for High-Dimensional Bayesian Optimization Using Synthetic Data for Machine Learning-based Childhood Vaccination Prediction in Narok, Kenya Delve into the Applicability of Advanced Optimizers for Multi-Task Learning Bridging SFT and RL: Dynamic Policy Optimization for Robust Reasoning Multi-Agent Decision-Focused Learning via Value-Aware Sequential Communication Predictive Entropy Links Calibration and Paraphrase Sensitivity in Medical Vision-Language Models Efficient Hierarchical Implicit Flow Q-learning for Offline Goal-conditioned Reinforcement Learning Modality-Aware Zero-Shot Pruning and Sparse Attention for Efficient Multimodal Edge Inference The nextAI Solution to the NeurIPS 2023 LLM Efficiency Challenge Feature-Label Modal Alignment for Robust Partial Multi-Label Learning Integrated electro-optic attention nonlinearities for transformers Toward World Models for Epidemiology Tracing the Chain: Deep Learning for Stepping-Stone Intrusion Detection Batch Distillation Data for Developing Machine Learning Anomaly Detection Methods
Super Apriel: One Checkpoint, Many Speeds
SLAM Labs, · 2026-04-23 · via cs.LG updates on arXiv.org

View PDF HTML (experimental)

Abstract:We release Super Apriel, a 15B-parameter supernet in which every decoder layer provides four trained mixer choices -- Full Attention (FA), Sliding Window Attention (SWA), Kimi Delta Attention (KDA), and Gated DeltaNet (GDN). A placement selects one mixer per layer; placements can be switched between requests at serving time without reloading weights, enabling multiple speed presets from a single checkpoint. The shared checkpoint also enables speculative decoding without a separate draft model. The all-FA preset matches the Apriel 1.6 teacher on all reported benchmarks; recommended hybrid presets span $2.9\times$ to $10.7\times$ decode throughput at 96% to 77% quality retention, with throughput advantages that compound at longer context lengths. With four mixer types across 48 layers, the configuration space is vast. A surrogate that predicts placement quality from the per-layer mixer assignment makes the speed-quality landscape tractable and identifies the best tradeoffs at each speed level. We investigate whether the best configurations at each speed level can be identified early in training or only after convergence. Rankings stabilize quickly at 0.5B scale, but the most efficient configurations exhibit higher instability at 15B, cautioning against extrapolation from smaller models. Super Apriel is trained by stochastic distillation from a frozen Apriel 1.6 teacher, followed by supervised fine-tuning. We release the supernet weights, Fast-LLM training code, vLLM serving code, and a placement optimization toolkit.
Comments: Models: this https URL and this https URL . Dev model: this https URL . Training code: this https URL . Async RL: this https URL . Training logs: this https URL
Subjects: Machine Learning (cs.LG)
ACM classes: I.2.6; I.2.7
Cite as: arXiv:2604.19877 [cs.LG]
  (or arXiv:2604.19877v1 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2604.19877

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Torsten Scholak [view email]
[v1] Tue, 21 Apr 2026 18:00:25 UTC (7,594 KB)

此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。