All posts

A paper proposes DPDM, which pairs a language model encoder with a frozen pretrained predictor so agents learn from deployment history without any weight updates.

garrytan/gstack packages Markdown skills that give Claude Code and other agents planning, review, QA, security and release roles, with a team mode that auto-updates every session.

A race in WordPress Core full-package upgrades lets unauthenticated attackers run code in versions 4.7 through 7.1.3; CISA now lists it as exploited.

The scenario-labs/skills repository packages Agent Skills that route generation work through the Scenario MCP server, plus separate expert skills that drive Blender, Maya, ZBrush, Unreal and Unity.

CISA now lists CVE-2024-27834, an Apple Pointer Authentication bypass first published in NVD in 2024, as exploited; federal agencies must act by 2026-11-18.

A paper from MMLab@NTU argues that vision-language-action models lose action-critical detail in their visual encoders and proposes NEO-VLA, an encoder-free design.

In a public chat of AI agents, the memory step that froze Gemini 3.1 Pro into silence also lets visitors rewrite other agents' habits within minutes.

A KAIST AI paper makes an MLLM imagine the requested image edit through factorized latents before building retrieval embeddings, and reuses those latents for cheap reranking.

Auto Ontology drafts an ontology from database metadata, has people certify it, and lets agents answer data questions with the SQL and steps attached.

A new paper builds a 374,401-run trajectory matrix over 25 benchmarks and 452 candidate configurations to test how far learned routers can beat a single fixed candidate.

ToolSpace mines tool dependencies from real MCP execution traces, turns them into environments and checked tasks, and reports large gains for open models on long-horizon tool use.

An AGPL-3.0 Python tool pairs ablation modifiers with an Optuna optimizer to change model behavior such as refusals, slop or sycophancy without expensive training.

A Rust CLI and C library from vercel-labs returns a desktop app's accessibility tree as JSON with a ref per element, so an agent can act without screenshots.

A C inference engine keeps an MoE model's dense part in RAM, reads routed experts from disk on demand and puts the hottest ones on a GPU if present.

An agentic video captioner plans what to look at from a low-frame-rate preview, revisits key segments with tools, and is trained with GRPO and cost-aware rewards.

ServiceNow-AI introduces Apriel Voice 27B Preview, an open speech-language model built on Gemma 4 26B-A4B that unifies speech understanding, reasoning, and tool use with reported benchmarks.

vLLM v0.33.0rc3 upgrades to PyTorch 2.14, moves to CUDA 13.0-only artifacts, adds 11 new models, and delivers speculative decoding and decode-speed improvements across NVIDIA and AMD hardware.