·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
A Multi-Model Metric-based Selection Framework for Abstractive Text summarization2h◆Piercing Gilbreath's Conjecture: From Deep Number Theory Insights to Fintech and Cybersecurity2h◆Operator-on-F complements value-equivalence: a planning-time diagnostic for latent world models2h◆REDDIT: Correcting Model-Generated Timestamp Drift in ASR without Forgetting via Replay-Based Distribution Editing2h◆AnchorPrune: Relevance-Anchored Contextual Expansion for Visual Token Pruning2h◆HELP: Human-Efficient Large-Scale Robot Post-Training with Rollout Segmentation2h◆AgentCheck: A Reproduce-Intervene-Mitigate Workbench for LLM Agents over MCP2h◆An Explainable Agentic System for Detection of Conversational Scams with Summary-Based Memory2h◆Introducing Human-Centeredness in AI-Assisted Lexicography2h◆Generalized Distribution-Free Semi-Supervised Learning with Risk Rewrite2h◆Removable Defects: The Economics and Limits of Deliberate Deficiency2h◆Are we Merging the Right Models? Impact of Expert Training Duration on Model Merging for LLMs2h◆Mind the Gap: Promises and Pitfalls of Hierarchical Planning in LeWorldModel2h◆Hallo4D: Multi-Modal Hallucination Mitigation for Consistent Spatio-Temporal Generation2h◆The Capacity of Thought: Benchmarking Llama 3.2 in Semantic fMRI Neural Language Decoding and Improving the Huth Encoding-Model Baseline2h◆Token Reduction Is Not Cost Reduction2h◆FinResearchBench II: A Deep Research Benchmark with Consensus-Derived Gold Rubrics for Distinguishing Financial Report Quality2h◆WikiSTAR: A System for Shedding Light on the Hidden History of Scientific Wikipedia Articles2h◆KnowAct-GUIClaw: Know Deeply, Act Perfectly, Personal GUI Assistant with Self-Evolving Memory and Skill2h◆Can LLMs Write Reliable Rubrics? A Meta-Evaluation for Experiment Reproduction2h◆A Multi-Model Metric-based Selection Framework for Abstractive Text summarization2h◆Piercing Gilbreath's Conjecture: From Deep Number Theory Insights to Fintech and Cybersecurity2h◆Operator-on-F complements value-equivalence: a planning-time diagnostic for latent world models2h◆REDDIT: Correcting Model-Generated Timestamp Drift in ASR without Forgetting via Replay-Based Distribution Editing2h◆AnchorPrune: Relevance-Anchored Contextual Expansion for Visual Token Pruning2h◆HELP: Human-Efficient Large-Scale Robot Post-Training with Rollout Segmentation2h◆AgentCheck: A Reproduce-Intervene-Mitigate Workbench for LLM Agents over MCP2h◆An Explainable Agentic System for Detection of Conversational Scams with Summary-Based Memory2h◆Introducing Human-Centeredness in AI-Assisted Lexicography2h◆Generalized Distribution-Free Semi-Supervised Learning with Risk Rewrite2h◆Removable Defects: The Economics and Limits of Deliberate Deficiency2h◆Are we Merging the Right Models? Impact of Expert Training Duration on Model Merging for LLMs2h◆Mind the Gap: Promises and Pitfalls of Hierarchical Planning in LeWorldModel2h◆Hallo4D: Multi-Modal Hallucination Mitigation for Consistent Spatio-Temporal Generation2h◆The Capacity of Thought: Benchmarking Llama 3.2 in Semantic fMRI Neural Language Decoding and Improving the Huth Encoding-Model Baseline2h◆Token Reduction Is Not Cost Reduction2h◆FinResearchBench II: A Deep Research Benchmark with Consensus-Derived Gold Rubrics for Distinguishing Financial Report Quality2h◆WikiSTAR: A System for Shedding Light on the Hidden History of Scientific Wikipedia Articles2h◆KnowAct-GUIClaw: Know Deeply, Act Perfectly, Personal GUI Assistant with Self-Evolving Memory and Skill2h◆Can LLMs Write Reliable Rubrics? A Meta-Evaluation for Experiment Reproduction2h◆
News/Repeated Shared Access Enables Grokking, but Edit Propagation Depends on an Addressable Memory
arxiv
PublishedJune 25, 2026 at 4:00 AM
—neutral

Repeated Shared Access Enables Grokking, but Edit Propagation Depends on an Addressable Memory

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

arXiv:2606.20737v2 Announce Type: replace Abstract: We study factual edit propagation in a controlled synthetic knowledge-graph QA setting using a 2x2 grid that crosses loop recurrence with shared-memory access: a dense transformer (Dense), a looped transformer (Loop), a dense backbone with shared m

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
arxivA Multi-Model Metric-based Selection Framework for Abstractive Text summarization2harxivPiercing Gilbreath's Conjecture: From Deep Number Theory Insights to Fintech and Cybersecurity2harxivOperator-on-F complements value-equivalence: a planning-time diagnostic for latent world models2harxivREDDIT: Correcting Model-Generated Timestamp Drift in ASR without Forgetting via Replay-Based Distribution Editing2h
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