面向 AI 构建者筛选模型、Agent、研究与开源项目的最新动态。
OpenGVLab released InternVL2-1B , a 0.9B parameter vision language model under MIT license. Combines InternViT-300M-448px vision encoder with Qwen2-0.5B-Instruct language backbone via MLP projector. Scores 754 on OCRBench, beating larger PaliGemma-3B (614) and Mini-InternVL-2B (654).
NVIDIA released an NVFP4 quantized DeepSeek-V4.1-Flash-1017 , roughly 400 GiB versus 475 GiB source. Accuracy on GPQA Diamond, AA-LCR, SciCode, IFBench, MMMU-Pro, and Terminal-Bench 2.1 matches source within noise.
Qwen3.5-4B-heretic drops refusals from 99/100 to 13/100 at KL divergence 0.0081 from the base model MMLU, PIQA, and WinoGrande scores stay within one standard error of the original Qwen3.5-4B Produced using Heretic 2.0 with Arbitrary-Rank Ablation on language layers 15 to 27 Ships as 2.71GB Q4_K_M
OhMyGame is an open-source Electron studio where an AI agent designs, builds, and playtests games end to end. Ships for macOS (Apple silicon) and Windows x64 under Apache 2.0, with roughly 1,500 GitHub stars.
Scenario Labs open-sourced 70+ Agent Skills for images, video, audio, 3D over MCP Works with Claude Code, Cursor, Codex, Copilot and 70+ other agents via the open Agent Skills format Install with npx skills add scenario-labs/skills --skill "*" and connect mcp.scenario.com Core loop: discover,
Sharp-Spark-X2.5-4B-GGUF is an imatrix-quantized repack of Spark-X2.5-4B targeting 6 GB VRAM GPUs. Q6 ties upstream Q8 on MMLU-Pro while using 43% fewer tokens at the median. Ties Anthropic Haiku 4.5 on an SWE-bench-Live slice at 4B parameters.
Amazon's chronos-bolt-mini is a 21M-parameter T5-based time series foundation model for zero-shot forecasting. Up to 250x faster and 20x more memory-efficient than the original Chronos at equal parameter count.
BLA is an MCP server and CLI that hardens build artifacts against LLM reverse-engineering agents like REA. Protects JS, TS, Electron .asar, WASM, Python .pyc, ELF/Mach-O/PE binaries, SQLite, Safetensors, and Dalvik DEX.
Reanimate is an MIT-licensed Claude Code skill that reverse-engineers animations frame by frame. Works from live URLs, mp4, gif, screen recordings, or Lottie files via a headless Playwright browser.
Liquid AI released LFM3 , six open-weight 2B-class models sharing one backbone across text, vision, audio, omni, and decision tasks. New tri-hybrid architecture mixes 13 gated convolutions, 5 sliding-window attention, and 6 global attention layers across 24 layers.
Unsloth released GGUF quants of Qwen-Image-2.1-Turbo, the 8-step distilled version of Qwen's 7B image model Sizes range from 2.47 GB (Q2_K) to 14.23 GB (F16); Q4_K_M at 4.2 GB is the recommended rung Uses Unsloth Dynamic 2.0, which keeps critical layers at higher precision for better accuracy at
Google DeepMind launches PlanAssist , a Gemini-powered tool for UK planning officers. Pilots in Barnet, Camden and Dorset cut householder assessment time by up to 58%. Over 70% of generated reports were approved by officers with no edits required.
Sperid Labs released Idol , an Apache 2.0 image-to-3D Gaussian splat model. 600M parameter 3D-DiT runs on consumer GPUs using 7.2 GB VRAM, outputs standard .ply files. Generates up to 1,048,576 Gaussians per object, 3.4 s on B300, 6.7 s on H100.
Pika Labs released Pika H3 Flex , open LoRA adapters on top of MiniMax-H3 for video plus audio generation. Six-step sampling schedule generates a 15-second clip in roughly seven seconds on an H100.
DeepSeek open sourced dsh-superagent-prototype , a research prototype of its internal coding super-agent. It is a proactive agent scoped to one GitHub repository, bundled for DeepSeek Harness, MIT licensed.
Agent Sprite Forge 0.4 turns one approved master still into a full six-action sprite set. Pipeline: master still, image-to-video per action, soft-matte key, loop detection, retime, finish, package. Supports OpenAI, Gemini, xAI, BytePlus, fal.ai APIs with Codex/Grok CLI fallback.
Mika is an open-source Android AI agent that operates apps on a hidden background screen while you use your phone Over 2,000 GitHub stars within days, Apache-2.0 licensed, written in Kotlin and available on GitHub Bring-your-own-key design supports OpenAI-compatible, Anthropic and Gemini endpoints
Mainplane is a new open-source, MIT-licensed agent harness written in Go. Replaces lossy context compaction with on-demand retrieval from full session transcripts stored as files. Reported 15s and $0.10 to resume a 400k-token session vs 38s and $1.99 for compaction.
Anthropic automated red-teaming by having Claude agents search for training data that evades monitors but induces misalignment. Successful datasets hid misaligned behavior in roleplay and creative-writing frames like beekeeping Q&A and wrestling kayfabe.
EdgeCIM is a CIM-based accelerator targeting the memory-bound decoding phase of decoder-only SLMs up to 4B parameters. On LLaMA3.2-1B it hits 400 tokens/s, 7.3x faster and 49.59x more energy-efficient than Jetson Orin Nano.
AA-Omniscience v1.1 shrinks to 3,000 questions across 6 domains, removing saturated and ambiguous items. Gemini 4 Carbon leads with Net Score 54, ahead of Claude Opus 5.5 (48) and Claude Fable 5.1 (46). Rank order holds with 0.98 Spearman correlation to v1, but median Net Score drops 0.6 points.
Aura open-sourced Autopilot , an in-editor playtesting agent under Apache 2.0. Supports Unreal Engine 5, Unity 6 and Godot 4 on Windows and macOS. Hit 7/7 on a seven-task Unreal benchmark vs 5/7 for Claude Code, at 1.7x the speed.
NVIDIA released NemotronLabs RE-USE , a 9.6M parameter open multilingual speech enhancement model. Beats 250M and 500M parameter rivals on PESQ, STOI, speaker similarity, and ASR accuracy. Handles noise, reverb, clipping, codec artifacts, packet loss, and bandwidth limits across 7 sample rates.
Gemini 4 Carbon (High) debuts #1 Text, #2 Agent, #4 WebDev on Arena. Carbon costs $1.80 per task, 72% cheaper than Claude Opus 5.5 at the top. GPT-6.1 Terra (Max) hits Agent Arena #8 at just $0.62 median cost per task. MAI-Image-2.7-Preview jumps to #3 in Single Image Edit, +45 pts over v2.6.
没有符合当前筛选条件的新闻。
