·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
Neocloud Lambda secures $1B in debt to buy more chips2h◆An Anthropic researcher just gave us a peek at self-improving AI3h◆Open-weight AI companies are the Valley’s hottest acquisition targets4h◆Trump’s EPA wants to let data centers hide their air pollution6h◆Anthropic gets its first court win over the Pentagon’s supply-chain risk label10h◆Meta executive leaves for OpenAI as the social media giant faces growing scrutiny in India10h◆Fine-Tuning of Transformer models with Frames19h◆Feature Transformation Enhanced Jacobi Polynomial Graph Filtering for Graph Anomaly Detection19h◆Naive Prompt Optimization: Rethinking the Need for Complex Prompt Search19h◆TRACE-CRC: Trajectory-Adaptive Conformal Risk Control for Multi-Step Channel State Information Prediction19h◆Syntax vs. Semantics: How Transformers Learn Deep Dependencies19h◆NeuronFuzz: Safety Neuron Guided Fuzzing for LLM Safety Evaluation19h◆Data-driven Koopman mode approximation: A neural power iteration algorithm19h◆FaulT-Bench: Towards Benchmarking Network Troubleshooting LLM Agents under Unreliable User Tickets19h◆You Don't Need to Run Every Eval19h◆Unsupervised Post-Training of Foundation Models: A Survey19h◆Drift-Adaptive ICU Intervention Prediction: Freezing the Physiological Encoder for Auditable Model Updating19h◆Jiuge-Tuiqiao: An Interpretable Human-AI System for Classical Chinese Poetry Refinement19h◆Provable one-poison backdoor attacks on linear models and ReLU neural networks19h◆Is Your Neighborhood Safe? Place-based Stigma in Large Language Models' Urban Safety Judgments19h◆Neocloud Lambda secures $1B in debt to buy more chips2h◆An Anthropic researcher just gave us a peek at self-improving AI3h◆Open-weight AI companies are the Valley’s hottest acquisition targets4h◆Trump’s EPA wants to let data centers hide their air pollution6h◆Anthropic gets its first court win over the Pentagon’s supply-chain risk label10h◆Meta executive leaves for OpenAI as the social media giant faces growing scrutiny in India10h◆Fine-Tuning of Transformer models with Frames19h◆Feature Transformation Enhanced Jacobi Polynomial Graph Filtering for Graph Anomaly Detection19h◆Naive Prompt Optimization: Rethinking the Need for Complex Prompt Search19h◆TRACE-CRC: Trajectory-Adaptive Conformal Risk Control for Multi-Step Channel State Information Prediction19h◆Syntax vs. Semantics: How Transformers Learn Deep Dependencies19h◆NeuronFuzz: Safety Neuron Guided Fuzzing for LLM Safety Evaluation19h◆Data-driven Koopman mode approximation: A neural power iteration algorithm19h◆FaulT-Bench: Towards Benchmarking Network Troubleshooting LLM Agents under Unreliable User Tickets19h◆You Don't Need to Run Every Eval19h◆Unsupervised Post-Training of Foundation Models: A Survey19h◆Drift-Adaptive ICU Intervention Prediction: Freezing the Physiological Encoder for Auditable Model Updating19h◆Jiuge-Tuiqiao: An Interpretable Human-AI System for Classical Chinese Poetry Refinement19h◆Provable one-poison backdoor attacks on linear models and ReLU neural networks19h◆Is Your Neighborhood Safe? Place-based Stigma in Large Language Models' Urban Safety Judgments19h◆
News/Ternary Mamba: Grouped Quantization-Aware Training of W1.58A16 State Space Models
arxiv
PublishedJune 17, 2026 at 4:00 AM
—neutral

Ternary Mamba: Grouped Quantization-Aware Training of W1.58A16 State Space Models

Source
arxiv.orgfull article ↗
Read on arxiv→
Publisher summary· verbatim

arXiv:2606.18114v1 Announce Type: cross Abstract: State Space Models (SSMs) such as Mamba-2 offer linear-time inference but their memory footprint limits edge deployment. Prior ternary SSM work (Slender-Mamba) trains from scratch on 150B tokens; we show a pretrained checkpoint suffices, reducing the

Stay posted· Newsletter

A 5-min weekly brief — top movers, price watch, story of the week.

// no spam · unsubscribe one-click · free forever

Discussion
Source
↗
arxiv
Read original ↗All from arxiv →

No replies yet. Be first.

Source
↗
arxiv
Read original ↗All from arxiv →

Related coverage

More from ARXIV
arxivFine-Tuning of Transformer models with Frames19harxivFeature Transformation Enhanced Jacobi Polynomial Graph Filtering for Graph Anomaly Detection19harxivNaive Prompt Optimization: Rethinking the Need for Complex Prompt Search19harxivTRACE-CRC: Trajectory-Adaptive Conformal Risk Control for Multi-Step Channel State Information Prediction19h
The Bubble Brief
WEEKLY

Read AI insights every Tuesday — top movers, new releases, story of the week.

// no spam · unsubscribe one-click · free forever

Originally published on arxiv ↗
HomeModelsNews