·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
Beyond a Single Direction: Chain-of-Thought Disrupts Simple Steering of Refusal1h◆RobustSpeechFlow: Learning Robust Text-to-Speech Trajectories via Augmentation-based Contrastive Flow Matching1h◆Causal-Audit: Explicit and Auditable Graph-based Reasoning via Target-Aware Causal Chain Construction1h◆Behavioral Controllability of Agentic Models for Information Extraction: From Fixed Workflows to Reflective Agents1h◆NeurOWL: An LLM-Based Neural-symbolic Framework for Incomplete OWL Ontology Reasoning1h◆AgentFAIR: A Multi-Agent Collaborative Framework for FAIRness Evaluation of Geospatial Datasets1h◆Knowledge-Centric Agents for Workflow Generation1h◆DSWorld: A Data Science World Model for Efficient Autonomous Agents1h◆A Formally Grounded ODRL Evaluator: Implementation and Comparison1h◆Closing the AI Trust Gap: The Case for Independent Certification for Trustworthy AI1h◆SciForge: An AI-Native, Multimodal Workbench for Scientific Discovery1h◆Harmonizing AI Safety Thresholds1h◆CRAFT: Clustering Rubrics to Diagnose Weak LLM Capabilities and Generate Targeted Fine-Tuning Data1h◆Empathy as Predictive Misalignment Tolerance: A Co-Regulation Framework and the Regime Structure of Dialogue Repair1h◆How Does Empowering Users with Greater System Control Affect News Filter Bubbles?1h◆Structure of the Circular-Dyadic Convolution Error1h◆AV-JEPA: Extending LeJEPA to Audio-Visual Self-Supervised Learning1h◆Data-driven Video Codec with Implicit Neural Representations1h◆Lazy Arithmetic using Systolic Arrays for Closing the Verification Gap on Embedded Systems1h◆Verbalizable Representations Form a Global Workspace in Language Models1h◆Beyond a Single Direction: Chain-of-Thought Disrupts Simple Steering of Refusal1h◆RobustSpeechFlow: Learning Robust Text-to-Speech Trajectories via Augmentation-based Contrastive Flow Matching1h◆Causal-Audit: Explicit and Auditable Graph-based Reasoning via Target-Aware Causal Chain Construction1h◆Behavioral Controllability of Agentic Models for Information Extraction: From Fixed Workflows to Reflective Agents1h◆NeurOWL: An LLM-Based Neural-symbolic Framework for Incomplete OWL Ontology Reasoning1h◆AgentFAIR: A Multi-Agent Collaborative Framework for FAIRness Evaluation of Geospatial Datasets1h◆Knowledge-Centric Agents for Workflow Generation1h◆DSWorld: A Data Science World Model for Efficient Autonomous Agents1h◆A Formally Grounded ODRL Evaluator: Implementation and Comparison1h◆Closing the AI Trust Gap: The Case for Independent Certification for Trustworthy AI1h◆SciForge: An AI-Native, Multimodal Workbench for Scientific Discovery1h◆Harmonizing AI Safety Thresholds1h◆CRAFT: Clustering Rubrics to Diagnose Weak LLM Capabilities and Generate Targeted Fine-Tuning Data1h◆Empathy as Predictive Misalignment Tolerance: A Co-Regulation Framework and the Regime Structure of Dialogue Repair1h◆How Does Empowering Users with Greater System Control Affect News Filter Bubbles?1h◆Structure of the Circular-Dyadic Convolution Error1h◆AV-JEPA: Extending LeJEPA to Audio-Visual Self-Supervised Learning1h◆Data-driven Video Codec with Implicit Neural Representations1h◆Lazy Arithmetic using Systolic Arrays for Closing the Verification Gap on Embedded Systems1h◆Verbalizable Representations Form a Global Workspace in Language Models1h◆
News/REAL: REtrieval-reAsoning and Logic-constructed Attention Behaviors for Long-Context KV Cache Compression
arxiv
PublishedJuly 13, 2026 at 4:00 AM
—neutral

REAL: REtrieval-reAsoning and Logic-constructed Attention Behaviors for Long-Context KV Cache Compression

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

arXiv:2508.15806v2 Announce Type: replace-cross Abstract: The growing sequence length of large language models poses significant challenges for key-value (KV) caches. Existing state-of-the-art cache eviction methods primarily analyze the inference behavior of attention heads in successful retrieval-

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
arxivBeyond a Single Direction: Chain-of-Thought Disrupts Simple Steering of Refusal1harxivRobustSpeechFlow: Learning Robust Text-to-Speech Trajectories via Augmentation-based Contrastive Flow Matching1harxivCausal-Audit: Explicit and Auditable Graph-based Reasoning via Target-Aware Causal Chain Construction1harxivBehavioral Controllability of Agentic Models for Information Extraction: From Fixed Workflows to Reflective Agents1h
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