groundy

Groundy — independent coverage of developer tools, infrastructure, and platforms

agents

Can Self-Evolving AI Agents Drift Without a Human in the Loop?

Self-evolving AI agents drift without checkpoints: 94% of reviewers miss agent sabotage, safety hardening does not transfer across domains, and stale memory degrades tasks.

6min · 18:38 · jun 06
culture

A Covert LLM Persuasion Experiment Was Shut Down: How Far Did the Bots Get?

A 2026 analysis of the bot comment archive from a halted Reddit experiment catalogs fabricated identities and bias triggers, but early shutdown leaves harm unmeasurable.

6min · 15:42 · jun 06
infra

Indexing Images for RAG: kapa.ai's Approach to Multimodal Retrieval

kapa.ai's data shows indexing image captions at ingestion adds 1-6% query overhead versus 27-51% for raw query-time vision, shifting recall risk to caption fidelity.

models

Can LLMs Leak Training Data? A New Test Splits Capacity From Intent

PropMe splits memorization audits into capability and propensity, showing that single-metric leakage reports understate what targeted prompts can extract from LLMs.

policy

GDPR Rectification Rights Have No Clear Owner in ML Supply Chains

A 2026 arXiv paper shows GDPR rectification and erasure rights become unenforceable across ML supply chains where no party can trace a subject's data inside trained weights.

security

Benchmarking RAG Over Cyber Threat Intelligence: Where Retrieval Breaks

CTIConnect, a KDD 2026 benchmark of 1,860 QA pairs across five CTI feeds, shows retrieval quality, not model size, determines copilot accuracy across ten LLMs.

models

When an AI Agent's Tools Break, Can It Recover? A New Benchmark

ToolMaze, a new arXiv benchmark, shows LLM agents' recovery rates drop 37% when tools return corrupted data, exposing a gap in how agent reliability is measured.

industry

US Hyperscale Data Centers: A Carbon Audit That Recasts AI Power Costs

A facility-level audit of 403 US hyperscale centers finds 545 gCO2/kWh, 48% above the grid average. Siting in fossil-heavy regions, not PPAs, determines actual emissions.

featured editor's picks
most popular last 6 days · by traffic
top highest rated
feed all 419 articles · reverse chronological
  1. jun 06 agents Can Self-Evolving AI Agents Drift Without a Human in the Loop?
  2. jun 06 culture A Covert LLM Persuasion Experiment Was Shut Down: How Far Did the Bots Get?
  3. jun 06 infra Indexing Images for RAG: kapa.ai's Approach to Multimodal Retrieval
  4. jun 06 models Can LLMs Leak Training Data? A New Test Splits Capacity From Intent
  5. jun 06 policy GDPR Rectification Rights Have No Clear Owner in ML Supply Chains
  6. jun 06 security Benchmarking RAG Over Cyber Threat Intelligence: Where Retrieval Breaks
  7. jun 06 models When an AI Agent's Tools Break, Can It Recover? A New Benchmark
  8. jun 06 industry US Hyperscale Data Centers: A Carbon Audit That Recasts AI Power Costs
  9. jun 05 infra The RTX Spark Bet on Unified Memory for Local LLMs: Where Bandwidth Caps It
  10. jun 05 infra Reading Vercel's Fluid Compute vs Cloudflare Workers Benchmark
  11. jun 05 agents Fine-Tuning Multi-Agent LLM Systems: RL Enters Where Prompt Tweaks Stall
  12. jun 05 security Stronger Safety Alignment Made LLMs Easier to Jailbreak, Not Harder
  13. jun 05 security SAML Signature Bypass Is Back: Inside the SAMLStorm Vulnerability Class
  14. jun 05 policy When LLM Safety Lives at Inference, Not Training: A Certification Gap
  15. jun 05 culture Do LLMs Understand Idioms in Low-Resource Languages?
  16. jun 05 infra Does CUDA Tile Match Hand-Tuned Kernels on Hopper and Blackwell?
  17. jun 05 security SAMLStorm: The SAML Signature Bug That Forges Valid SSO Logins
  18. jun 05 models MiniMax M3 Bets on Sparse Attention for 1M Context. Does the Math Hold?
  19. jun 05 models Can One Model Handle Every CAD Task? UniCAD Tests It
  20. jun 05 models Do Foundation Models Actually Learn Relational Structure In-Context?
  21. jun 05 models Can LLMs Write Better Research Paper Titles Than Authors?
  22. jun 05 models Does Information-Theoretic Example Selection Beat kNN for In-Context Learning?
  23. jun 05 infra Pod-Level Remote Attestation in Kubernetes: Confidential Workloads on dstack
  24. jun 05 models Do Concept Bottleneck Model Benchmarks Measure Interpretability or Dataset Bias?
  25. jun 05 agents Cascading Hallucination in Agentic RAG: When One Bad Retrieval Poisons the Chain
  26. jun 05 security Vercel's Flags SDK Exposed Feature-Flag Definitions via CVE-2025-46332
  27. jun 05 models Continuous Bit-Width Quantization vs Fixed INT4: Does LiftQuant Beat Discrete?
  28. jun 04 models Federated Learning for Industrial IoT Anomaly Detection: The Data-Locality Tradeoff
  29. jun 04 infra Generating GPU Kernels for Moore Threads Silicon: Can LLMs Break CUDA Lock-In?
  30. jun 04 devtools Alibaba's Open Code Review Moves AI Review Into the CLI, Not the PR
  31. jun 04 infra Microsoft's Azure Linux Goes General-Purpose: The Container Base-Image Play
  32. jun 04 models Reading Failed LLM Reasoning Traces Won't Tell You Which Ones RL Can Fix
  33. jun 04 agents Can AI Agents Build Other Agents? The Meta-Agent Challenge Says Mostly Not Yet
  34. jun 04 models Can You Stitch Two Foundation Models Together Without Retraining?
  35. jun 04 infra Cloudflare Acquires VoidZero, the Company Behind Vite's Rust Toolchain
  36. jun 04 security Jailbreak Suffixes Hit Harder at Specific Token Positions, New GCG Variant Shows
  37. jun 04 policy When Should an LLM Forget You? A Benchmark for Deciding What Memory to Drop
  38. jun 04 security OpenAI Adds Lockdown Mode to ChatGPT, Shifting Prompt-Injection Risk to Users
  39. jun 04 policy When RL Training Rewards Capability-Seeking: A New Alignment Risk
  40. jun 04 models Do Reasoning LLMs Waste Tokens? OckBench Tries to Measure It
  41. jun 04 security Activation Steering Was Sold as LLM Control. New Work Makes It an Attack Surface
  42. jun 04 culture Can Teaching Logical Fallacies Inoculate People Against AI Misinformation?
  43. jun 04 devtools Vercel Ships Experimental Native CLI Binaries to Cut the Node Startup Tax
  44. jun 04 security Catching LLM Agents Leaking Credentials From Their Own Activations
  45. jun 04 policy Refusal Steering Targets Individual Experts in MoE LLMs
  46. jun 04 infra Putting a Datacenter V100 in a Gaming PC: The Local LLM Math
  47. jun 04 devtools Vercel Rebuilds Its Marketplace CLI for Agents Instead of Humans
  48. jun 04 security The 2026 npm Attacks Proved AI Coding Assistants Are a Supply-Chain Target
  49. jun 03 security ChatGPT's New Lockdown Mode Borrows Apple's Name for a Prompt-Injection Kill Switch
  50. jun 03 agents When MCP Tool Descriptions Don't Match the Code, Agents Trust the Lie
load older →