The Archive

AID-Guard introduces stateful authorization protocol for tool-using agents to prevent duplicate effects and state conflicts.

Graph engineering framework enabling LLM agents to coordinate heterogeneous expertise and parallel subtasks via system-level orchestration.

A frontier language model is only one component of an AI agent. The surrounding agent system—often called a harness—determines how the model receives... A frontier language model is only one component of an AI agent. The surrounding agent system—often called a harness—determines how the model receives context, uses tools, maintains state, responds to feedback, recovers from failure, and sustains progress over long-running tasks. The challenge is how to build the agent architecture that makes frontier language models work reliably on extended… Source

As AI agents become more capable and operate over longer horizons, building security and trust into the applications they power becomes increasingly important.... As AI agents become more capable and operate over longer horizons, building security and trust into the applications they power becomes increasingly important. Drawing on work with NVIDIA OpenShell, agent developers, open-source projects, and partners across the ecosystem, AI safety and security teams at NVIDIA offer their perspective on the emerging agent stack—including the role of each layer… Source

AI4AI-Bench isolates LLM agent capability to design training algorithms for recursive self-improvement, beyond hyperparameter tuning or data collection.

Study compares task-level vs. subtask-level skill induction in LLM agents; finds subtask-level code skills transfer more reliably across domains.

Proposes Software 3.0 paradigm shift: unified database, large model, and agent architecture as terminal form replacing instruction and data-driven models.

Empirical study of 557 coding agent sessions shows agents consult documentation differently than humans, challenging current technical documentation design.

“Runaway” AI, “rogue” agents, and “autonomous” actors—the current rhetoric would have you believe that AI agents are not only awake and aware, but angry at their creators. Prominent tech leaders such as Demis Hassabis, Dario Amodei, and Sam Altman push for regulation of these seemingly “superhuman” systems, while a separate faction, led by policy organizations…

LLM-based Multi-Agent Systems (MAS) achieve strong performance on complex reasoning tasks by coordinating multiple agents, but at the cost of substantial token consumption. Recent work on automatic topology design, ARG-Designer, has reframed this problem as autoregressive graph generation. However, its training objective provides no explicit incentive for the model to generate sparse and efficient topologies. We address this limitation by introducing a Reward-Guided Autoregressive Graph Generation (RGA-Designer) inspired by Reinforcement Learning from Human Feedback (RLHF). We train a reward ...

Slack is introducing dedicated channels where teams can vibe code together with AI agents instead of jumping between different tools and conversations. The Slack Code launch includes open, project-specific code channels with dedicated user tabs, alongside features that compare coding changes and preview HTML output before the project is shipped. "With Slack Code, when you have an idea or need to build a new feature, update a web page, or fix a bug, you simply tag in a coding agent like Anthropic's Claude or Cognition's Devin, and that agent then spins up a code channel to tackle the task," Sl...

Binance's Agent OS works with tools including ChatGPT, Claude Code, and Cursor.

Simon Willison discusses productivity metrics for AI coding agents, arguing lines of code can meaningfully measure agent output despite conventional skepticism.

NVIDIA Holoscan is a platform for building real-time AI applications at the edge, from medical imaging to robotics. HoloHub is its companion repository: a... NVIDIA Holoscan is a platform for building real-time AI applications at the edge, from medical imaging to robotics. HoloHub is its companion repository: a growing collection of reference applications and components that demonstrate what’s possible. We wanted to explore how a general-purpose coding agent could use the same examples, documentation, and development tools available to an engineer… Source

SPADE: self-play RL framework where LLMs generate adaptive training environments and learn from them, enabling continuous goal-distribution scaling for language agents.

Wasserstein entropic value-at-risk extends robust optimization for agents under uncertainty, enabling hedging against model misspecification beyond entropy balls.

LLM agents can execute training loops but lack strategy-level adaptation; empirical analysis shows post-training strategies remain locked despite experimental feedback.

AI agents are only as effective as the context they receive. Even with capable models and well-documented NVIDIA libraries, agents can spend extra steps finding... AI agents are only as effective as the context they receive. Even with capable models and well-documented NVIDIA libraries, agents can spend extra steps finding the right tools, burn tokens on dead ends, or struggle with specialized tasks. Skills package the instructions, examples, and tool guidance for agents to move faster from intent to solution. To measure whether these skills improve agent… Source

Harness Continual Learning framework enables agents to adapt prompts, memories, tools, and routing rules without retraining frozen models.

SkillForge: self-distillation framework enabling LLM agents to acquire project-specific knowledge for automated software issue resolution.

Atomistic simulation requires three things: knowledge of the science, compute-efficient implementation of simulations, and accessible interfaces to the... Atomistic simulation requires three things: knowledge of the science, compute-efficient implementation of simulations, and accessible interfaces to the simulation stack. The first remains the researcher’s domain, as no tool substitutes for knowing what to simulate or recognizing a physically meaningful result. NVIDIA ALCHEMI Toolkit, introduced earlier this year, has dramatically reduced the… Source

Comprehensive re-evaluation of memory-based self-improving agents reveals fragility: high variance across runs and sensitivity to task order undermine reliability claims.

Two large-scale surveys (N≈2,800 each) reveal asymmetric user acceptance of autonomous LLM agents in dating platforms: users accept delegation and reception differently.

StagedWorkspace versioned workspace system for knowledge-work agents ensures all parsed views, edits, and artifacts reference consistent workspace-state contracts.

Theoretical framework for LLM reward shaping in hybrid RL agents using bounded potential functions preserving optimal policy.

CABLE memory retrieval system for LLM agents addresses reachability of semantically-distant evidence in long-term conversational history via antecedent linking.

No GPUs, no Agents, just really, really, really good infra and distribution.

Humanoid robots hold great promise as general-purpose agents in human-centered environments, yet generalist vision-language-action (VLA) foundation models are not readily applicable to humanoid whole-body loco-manipulation. The high dimensionality and interdependence of humanoid motions make it challenging for conventional single-stage VLA architectures to coordinate locomotion, waist posture, and dual-arm manipulation effectively. Moreover, policies trained through offline behavior cloning can remain suboptimal during real-world deployment. Although online reinforcement learning can refine p...

Agents now write knowledge graphs, but knowledge-graph stores still carry defaults set when humans curated them: accept writes now and clean later, keep one time axis or none, treat every writer's facts as equally trustworthy, and leave governance to dashboards and middleware. These four defaults are individually convenient and jointly untenable under agent workloads. We present Quipu, an embeddable store that inverts all four: no fact enters except through a gate whose predicates evaluate the pending post-state; data, trust labels, verdicts, and the rules themselves are bitemporal; named gra...

Large Language Models (LLMs) have demonstrated capabilities in in-context learning, task decomposition, step-by-step reasoning, and code generation, driving their gradual evolution from text generation models into the core of agents capable of perceiving environments, invoking tools, and executing tasks. Traditional LLM Agents typically obtain information through webpages, documents, databases, or external tools and generate corresponding invocation sequences according to user goals; when this technology is further integrated with robotic systems, large language models begin to undertake func...