·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
Stochastic Sampling is Epistemically Shallow: The Dimensionality Gap Between Temperature Variation and Model Diversity in LLMs9h◆DecodeShare: Tracing the Shared Subspace of LLM Decode-Time Decisions9h◆PlanE: Meta Planning of Data, Tuning, and Inference for Extractive-based LLMs9h◆Incomplete Prompt Jailbreaks in Large Language Models9h◆Directional Hallucinations: Ideological Drift in News-Grounded LLM Question Answering9h◆Autonomous Topology Mutation: Safe Runtime Restructuring for Multi-Agent LLM Systems with Capability, State, and Shadow Invariants9h◆DFAH-Bench: Benchmarking Observable Agent Instability in Financial Decision-Making9h◆Isolating LLM Alignment from Regex: Zero Coverage and Metric-Dependent Divergence Under Adversarial Mutation9h◆AISE-Bench: A Full-Cycle Curated Benchmark for Information Seeking on Academic Knowledge Graphs9h◆MKEvolve: A Modular Multi-Agent Framework for Kernel Code Generation9h◆Telco-GAIA: Bilingual Benchmark for Agents in Telecom Domain9h◆SiGMA: Sign-Guided Merging and Adaptation for Multimodal Continual Instruction Tuning9h◆Attention Degradation, Function Token Anchoring, and the Limits of Attention-Based Intervention in Large Language Models9h◆StrideDiffusion: Accelerating Diffusion Models for Time-series Generation9h◆Enhancing Explainable Cardiac Diagnosis with Guide-Grounded Multimodal LLMs9h◆Efficient and Interpretable Body-Based Emotion Recognition with Lightweight Temporal Convolutional Networks9h◆Three-Pronged Spectral Control for Federated Parameter Efficient Fine Tuning9h◆Is Deep Research Reliable? Misleading Knowledge Induces False Conclusions9h◆Representing Entity Importance in AI Knowledge Systems: A Dual-Signal Framework of Audience Evaluation and Structural Authority9h◆V-DEAL: Diagnosing Video Safety De-Calibration as an Understanding-Refusal Coupling Failure9h◆Stochastic Sampling is Epistemically Shallow: The Dimensionality Gap Between Temperature Variation and Model Diversity in LLMs9h◆DecodeShare: Tracing the Shared Subspace of LLM Decode-Time Decisions9h◆PlanE: Meta Planning of Data, Tuning, and Inference for Extractive-based LLMs9h◆Incomplete Prompt Jailbreaks in Large Language Models9h◆Directional Hallucinations: Ideological Drift in News-Grounded LLM Question Answering9h◆Autonomous Topology Mutation: Safe Runtime Restructuring for Multi-Agent LLM Systems with Capability, State, and Shadow Invariants9h◆DFAH-Bench: Benchmarking Observable Agent Instability in Financial Decision-Making9h◆Isolating LLM Alignment from Regex: Zero Coverage and Metric-Dependent Divergence Under Adversarial Mutation9h◆AISE-Bench: A Full-Cycle Curated Benchmark for Information Seeking on Academic Knowledge Graphs9h◆MKEvolve: A Modular Multi-Agent Framework for Kernel Code Generation9h◆Telco-GAIA: Bilingual Benchmark for Agents in Telecom Domain9h◆SiGMA: Sign-Guided Merging and Adaptation for Multimodal Continual Instruction Tuning9h◆Attention Degradation, Function Token Anchoring, and the Limits of Attention-Based Intervention in Large Language Models9h◆StrideDiffusion: Accelerating Diffusion Models for Time-series Generation9h◆Enhancing Explainable Cardiac Diagnosis with Guide-Grounded Multimodal LLMs9h◆Efficient and Interpretable Body-Based Emotion Recognition with Lightweight Temporal Convolutional Networks9h◆Three-Pronged Spectral Control for Federated Parameter Efficient Fine Tuning9h◆Is Deep Research Reliable? Misleading Knowledge Induces False Conclusions9h◆Representing Entity Importance in AI Knowledge Systems: A Dual-Signal Framework of Audience Evaluation and Structural Authority9h◆V-DEAL: Diagnosing Video Safety De-Calibration as an Understanding-Refusal Coupling Failure9h◆
News/Error Certificates for KV-Cache Eviction via Randomized Design
arxiv
PublishedJuly 24, 2026 at 4:00 AM

Error Certificates for KV-Cache Eviction via Randomized Design

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

arXiv:2607.21475v1 Announce Type: cross Abstract: Deterministic KV-cache eviction keeps the top-$k$ tokens under an importance score and deletes the rest. We prove that this design cannot know what it destroyed: evicted values can be altered so that everything the serving system retains is unchanged

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
arxivStochastic Sampling is Epistemically Shallow: The Dimensionality Gap Between Temperature Variation and Model Diversity in LLMs9harxivDecodeShare: Tracing the Shared Subspace of LLM Decode-Time Decisions9harxivPlanE: Meta Planning of Data, Tuning, and Inference for Extractive-based LLMs9harxivIncomplete Prompt Jailbreaks in Large Language Models9h
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