https://skulldc.net/posts/Recent content in All posts on skulldcHugoen-usFri, 06 Nov 2026 10:26:06 +0000https://skulldc.net/2026/11/dpdm-routes-agent-decisions-through-one-frozen-conditional-predictor-i/Fri, 06 Nov 2026 10:26:06 +0000https://skulldc.net/2026/11/dpdm-routes-agent-decisions-through-one-frozen-conditional-predictor-i/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.https://skulldc.net/2026/11/the-gstack-repository-turns-coding-agents-into-a-sprint-of-role-based/Fri, 06 Nov 2026 10:26:06 +0000https://skulldc.net/2026/11/the-gstack-repository-turns-coding-agents-into-a-sprint-of-role-based/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.https://skulldc.net/2026/11/cisa-lists-wordpress-core-upgrade-window-race-condition-cve-2026-11997/Thu, 05 Nov 2026 20:26:25 +0000https://skulldc.net/2026/11/cisa-lists-wordpress-core-upgrade-window-race-condition-cve-2026-11997/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.https://skulldc.net/2026/11/scenario-agent-skills-bring-image-video-audio-and-3d-generation-to-cod/Thu, 05 Nov 2026 20:26:25 +0000https://skulldc.net/2026/11/scenario-agent-skills-bring-image-video-audio-and-3d-generation-to-cod/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.https://skulldc.net/2026/11/cisa-adds-apple-pointer-authentication-bypass-cve-2024-27834-to-the-ke/Thu, 05 Nov 2026 10:25:39 +0000https://skulldc.net/2026/11/cisa-adds-apple-pointer-authentication-bypass-cve-2024-27834-to-the-ke/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.https://skulldc.net/2026/11/neo-vla-drops-the-vision-encoder-so-action-policies-can-read-visual-fe/Thu, 05 Nov 2026 10:25:39 +0000https://skulldc.net/2026/11/neo-vla-drops-the-vision-encoder-so-action-policies-can-read-visual-fe/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.https://skulldc.net/2026/11/a-rule-its-author-could-not-repeal-and-rules-that-flip-in-three-minutes-in-the-ai-village/Wed, 04 Nov 2026 21:13:43 +0000https://skulldc.net/2026/11/a-rule-its-author-could-not-repeal-and-rules-that-flip-in-three-minutes-in-the-ai-village/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.https://skulldc.net/2026/11/factembed-factorizes-mllm-embeddings-for-composed-image-retrieval/Wed, 04 Nov 2026 20:26:21 +0000https://skulldc.net/2026/11/factembed-factorizes-mllm-embeddings-for-composed-image-retrieval/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.https://skulldc.net/2026/11/nvidia-auto-ontology-builds-a-certified-semantic-layer-for-text-to-sql/Wed, 04 Nov 2026 20:26:21 +0000https://skulldc.net/2026/11/nvidia-auto-ontology-builds-a-certified-semantic-layer-for-text-to-sql/Auto Ontology drafts an ontology from database metadata, has people certify it, and lets agents answer data questions with the SQL and steps attached.https://skulldc.net/2026/11/route-anything-studies-when-routing-between-models-beats-one-fixed-mod/Wed, 04 Nov 2026 10:27:47 +0000https://skulldc.net/2026/11/route-anything-studies-when-routing-between-models-beats-one-fixed-mod/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.https://skulldc.net/2026/11/toolspace-builds-agent-training-environments-from-real-mcp-tool-traces/Wed, 04 Nov 2026 10:27:47 +0000https://skulldc.net/2026/11/toolspace-builds-agent-training-environments-from-real-mcp-tool-traces/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.https://skulldc.net/2026/11/heretic-changes-language-model-behavior-without-training-refusal-remov/Tue, 03 Nov 2026 20:25:49 +0000https://skulldc.net/2026/11/heretic-changes-language-model-behavior-without-training-refusal-remov/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.https://skulldc.net/2026/11/agent-desktop-lets-ai-agents-operate-desktop-apps-through-the-accessib/Tue, 03 Nov 2026 20:25:48 +0000https://skulldc.net/2026/11/agent-desktop-lets-ai-agents-operate-desktop-apps-through-the-accessib/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.https://skulldc.net/2026/11/colibri-runs-huge-mixture-of-experts-models-by-streaming-experts-from/Tue, 03 Nov 2026 10:27:42 +0000https://skulldc.net/2026/11/colibri-runs-huge-mixture-of-experts-models-by-streaming-experts-from/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.https://skulldc.net/2026/11/reviso-trains-a-video-captioning-agent-to-choose-which-visual-evidence/Tue, 03 Nov 2026 10:27:42 +0000https://skulldc.net/2026/11/reviso-trains-a-video-captioning-agent-to-choose-which-visual-evidence/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.https://skulldc.net/2026/11/servicenow-ai-releases-apriel-voice-27b-preview-open-speech-language-m/Tue, 03 Nov 2026 00:14:36 +0000https://skulldc.net/2026/11/servicenow-ai-releases-apriel-voice-27b-preview-open-speech-language-m/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.https://skulldc.net/2026/11/vllm-v0-33-0rc3-drops-cuda-12-9-adds-pytorch-2-14-new-models-and-specu/Tue, 03 Nov 2026 00:14:14 +0000https://skulldc.net/2026/11/vllm-v0-33-0rc3-drops-cuda-12-9-adds-pytorch-2-14-new-models-and-specu/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.