·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
Jensen Huang and the decline of the visionary showman founder CEO1h◆Mach2: Scientific Discovery Platform with Deterministic Execution Constraints8h◆OverServe: Resource Amplification Attacks on SaaS Agents via Tool--Prompt Co-Design8h◆CFGComposer: Compositional Audio-Text-to-Motion Generation via Copula Score Correction8h◆Multi-View 3D Reconstruction Using V-JEPA Features8h◆GingivitisSeg: A Public Dataset and Comparative Benchmark for Gingivitis Segmentation in Intraoral Images8h◆Can Model Answer Probabilities Represent Survey Populations? Conditional Prediction and Joint Validity in Jev Survey Simulation8h◆Contrastive Bootstrap Representation Learning for Label-Efficient Agricultural Image Classification8h◆SO-Arena: Benchmarking the Capability Spectrum and Reliability of System One Models8h◆The BEATS Dataset: Towards Understanding Music-Physiology Dynamics in Real-World Scenarios8h◆RESYNC: Residual Latent-State Synchronization for Deadline-Bound Streaming Inference8h◆LEAP: Supercharging LLMs for Formal Mathematics with Agentic Frameworks8h◆Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness8h◆Which Buildings Are Artificial Intelligence-Ready? A Measurement-Based Assessment Framework for AI Question Answering and Actuation8h◆Catch-Only-One: Non-Transferable Examples for Model-Specific Authorization8h◆DAISY: Fault-Guided Assertion Inference for Dafny8h◆Individualized Algorithmic Advice as a Strategic Signal on Competitive Markets8h◆Learning pharmacokinetic dynamics with structured neural ODEs for precision dosing8h◆ANCRe: Adaptive Neural Connection Reassignment for Efficient Depth Scaling8h◆Assessing Per-Sample Membership Inference Vulnerability without Retraining8h◆Jensen Huang and the decline of the visionary showman founder CEO1h◆Mach2: Scientific Discovery Platform with Deterministic Execution Constraints8h◆OverServe: Resource Amplification Attacks on SaaS Agents via Tool--Prompt Co-Design8h◆CFGComposer: Compositional Audio-Text-to-Motion Generation via Copula Score Correction8h◆Multi-View 3D Reconstruction Using V-JEPA Features8h◆GingivitisSeg: A Public Dataset and Comparative Benchmark for Gingivitis Segmentation in Intraoral Images8h◆Can Model Answer Probabilities Represent Survey Populations? Conditional Prediction and Joint Validity in Jev Survey Simulation8h◆Contrastive Bootstrap Representation Learning for Label-Efficient Agricultural Image Classification8h◆SO-Arena: Benchmarking the Capability Spectrum and Reliability of System One Models8h◆The BEATS Dataset: Towards Understanding Music-Physiology Dynamics in Real-World Scenarios8h◆RESYNC: Residual Latent-State Synchronization for Deadline-Bound Streaming Inference8h◆LEAP: Supercharging LLMs for Formal Mathematics with Agentic Frameworks8h◆Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness8h◆Which Buildings Are Artificial Intelligence-Ready? A Measurement-Based Assessment Framework for AI Question Answering and Actuation8h◆Catch-Only-One: Non-Transferable Examples for Model-Specific Authorization8h◆DAISY: Fault-Guided Assertion Inference for Dafny8h◆Individualized Algorithmic Advice as a Strategic Signal on Competitive Markets8h◆Learning pharmacokinetic dynamics with structured neural ODEs for precision dosing8h◆ANCRe: Adaptive Neural Connection Reassignment for Efficient Depth Scaling8h◆Assessing Per-Sample Membership Inference Vulnerability without Retraining8h◆
News/Cleaning an Agent's History Hurts When the World Still Holds the Error
arxiv
PublishedNovember 14, 2026 at 5:00 AM

Cleaning an Agent's History Hurts When the World Still Holds the Error

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

arXiv:2611.14654v1 Announce Type: new Abstract: Tool-using language-model agents accumulate histories of their own mistakes, and a system that compacts, repairs or hands off such a history decides what it says about those mistakes while their effects stay in the world. Prior work cleans the history

Stay posted· Newsletter

Three short emails a week: Monday, Wednesday and Friday. Top movers, new models and the most-referenced story.

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

Your email is used only to send The Bubble Brief. Privacy policy

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
arxivMach2: Scientific Discovery Platform with Deterministic Execution Constraints8harxivOverServe: Resource Amplification Attacks on SaaS Agents via Tool--Prompt Co-Design8harxivCFGComposer: Compositional Audio-Text-to-Motion Generation via Copula Score Correction8harxivMulti-View 3D Reconstruction Using V-JEPA Features8h
The Bubble Brief
MON · WED · FRI

Read AI insights every Monday, Wednesday and Friday — top movers, new models, story of the week.

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

Your email is used only to send The Bubble Brief. Privacy policy

Originally published on arxiv ↗
PrivacyTermsRefundsLegal notice|AboutMethodologyContact

© 2026 Databubble

Built by Marouane Gazouzi · Readyo, AI interview coach

HomeModelsNews