The Archive

Anthropic adds AGENTS.md support to Claude Code v2.1.277 for project instruction customization via modular architecture.

You’re deploying a model on a system. It starts up, prompts are getting responses. Now the hard question: Is this fast? Your instincts might lead you to send... You’re deploying a model on a system. It starts up, prompts are getting responses. Now the hard question: Is this fast? Your instincts might lead you to send curl commands, hand-roll an asyncio script, or vibe code yet another one-off load generator. All of these paths have the same problem: single-process performance limits, Python’s GIL capping concurrency, or numbers measured against a… Source

Jev, a new kind of AI model, is showing developers a cheaper and faster path to software intelligence.

Virginia Gov. Abigail Spanberger (D) ordered the state government to take steps that could empower local communities to have a larger say in data center development and slow down approvals in a state that is already home to the data center capital of the world. Executive Order 22 bans executive branch officials from signing non-disclosure agreements (NDAs) for data center projects, requires expedited noise regulations, and a review of backup-generation operations used by data centers, among other requirements. The order also establishes an AI task force responsible for evaluating how the stat...

Designer-RSI: agent framework with procedural memory learns reusable design skills from 230+ tools in professional graphics software.

The former CEO of Character.AI, which Disney previously sent a cease-and-desist letter to, will serve as the company's first-ever chief technology officer.

NLP model for occupational accident narratives shows cross-sector generalization challenges in automated accident-process role classification.

CodeMidas: RL pipeline scales agentic coding tasks by extracting diverse environments directly from open-source codebase source code.

Value-Sensitive Design analysis of 73K OpenClaw Reddit posts identifies 21 user values prioritized beyond task completion in agent delegation.

BrainWideBench: benchmark for evaluating transfer learning across animals and brain regions on multi-region neural recordings.

Multi-hop retrieval failures cluster predictably; dense-only ANN scoring weak vs LLM-judge pipelines for confidence calibration.

Benchmark comparing world models' continual learning on compositional tasks, measuring knowledge retention vs. speed of adaptation.

PCC+GCN hybrid framework uses particle competition dynamics for label refinement before GCN training under label noise.

Available Guardrails: selective prediction safety gate with certification of trustworthiness across reporting units and subgroups.

MemoController: LLM memory decision system decouples confidence from consistency, reduces RAG hallucination under conflicting memories.

λ-Controlled GRPO stabilizes reinforcement learning for flow-matching image generators by controlling importance ratio drift across denoising steps.

Gricea is an open-science platform for reproducible conversational AI research enabling study artifact sharing and reuse.

QuranicMMLU benchmark evaluates generative AI on Quranic Arabic across phonology, morphology, syntax, semantics, and pragmatics.

The Federal Register website briefly used an open source Chinese AI search tool.

A week after an Anthropic researcher’s doomsday warning rattled the AI world, the company’s CEO Dario Amodei has outlined his plan to “pace the frontier” of AI development. The proposal leans on independent safety evaluators and coordination between AI labs in democratic countries, and it’s already picked up some industry support, along with some pointed pushback from Nvidia’s Jensen Huang. Watch […]

COMPLEX provides certified lower-bound Lipschitz guarantees for multiparameter persistence module embeddings via closed-form slicing.

DiaVLo is a diagnostic framework for identifying behavior misalignments and influential concepts in vision-language models.

A week after an Anthropic researcher’s doomsday warning rattled the AI world, the company’s CEO Dario Amodei has outlined his plan to “pace the frontier” of AI development. The proposal leans on independent safety evaluators and coordination between AI labs in democratic countries, and it’s already picked up some industry support, along with some pointed pushback from Nvidia’s Jensen Huang. On […]

California Gov. Gavin Newsom (D) is positioning the state to take the lead on AI oversight, including the potential to mandate a "kill switch" for frontier models, with a new executive order issued Friday. Newsom's order directs the state to convene a group of experts that will deliver recommendations within two months on how to strengthen AI safety measures in state law. Newsom wants the group to consider how the state could require AI companies to embed independent verification groups onsite for regular audits, make their transparency reports and risk assessments subject to standards of ind...

Gated attention heads enable abstention and noise filtering missing from softmax attention in language model pretraining.

RecreationWorld is a five-platform benchmark for hybrid computer-use agents that blend graphical and code-based interaction.

Stratechery weekly digest covering tech industry trends including Salesforce strategy and regional perspectives, not AI-specific.

Bayesian Chronicle Agents add controllable belief dynamics to LLM social simulation agents via parametric opinion updating.

Probe of Internal Recognition detects concealed knowledge in LLMs using forensic methods adapted from psychology.

Machine learning improves critical heat flux prediction for nuclear reactor safety vs. traditional correlations.