Archive
新模型、開放權重、多模態、語音、影像與模型架構。
最新模型
The new benchmark uses 3,300 parallel questions in six languages and 5,340 French questions to test whether models can access the same everyday and long-tail knowledge across languages. Larger models are more accurate overall, but still fail to eliminate gaps caused by pop culture, recent events, and language switching.
AI 研究與訓練
A new study replaces the fixed number of reasoning trajectories generated for every problem with a strategy that concentrates additional compute on samples exhibiting greater reward disagreement. The authors report higher rollout efficiency on mathematics and coding tasks, but have not released their code, while actual GPU time and scalability to larger models remain to be verified.
最新模型
AMD has released six Instella-MoE checkpoints spanning pretraining, mid-training, long-context extension, SFT, DPO, and reinforcement learning, along with the training frameworks, data recipes, and inference code. The model uses Gated MLA and FarSkip-Collective to reduce attention and cross-GPU communication costs, but a formal technical report and independent evaluations are still pending.
最新模型
Induction Labs has introduced Photon-1, a 106B-A5B sparse MoE model that learns during pretraining solely by watching screen videos, without requiring mouse or keyboard action labels. The model first predicts the next latent screen state, then generates the actions needed to reach it, though current evidence of its performance comes primarily from the team’s internal evaluations.
代理工具
Cursor has transformed Auto mode into a Router that classifies each request and selects a model, allowing enterprises to switch among Intelligence, Balance, and Cost strategies. The company claims online testing shows savings of 60%, but hidden-by-default model selections and mandatory routing options for certain modes make auditability a key deployment concern.
最新模型
Preliminary testing by the UK AISI and the US CAISI shows that Kimi K3 scored 32% on ExploitBench, outperforming GLM-5.2 but trailing the tested US frontier models by a wide margin. The model also completed a 32-step simulated enterprise attack once, but the limited evaluation conditions and sample size do not support direct conclusions about real-world attack success rates.
AI 安全
Zenity has disclosed a patched CSRF-style attack chain: merely opening a crafted link while logged in could cause ChatGPT Agent Builder to automatically create, schedule, and run an attacker-controlled agent. The vulnerability did not bypass OAuth; instead, it inherited existing authorizations, highlighting a new trust gap among agent creation workflows, execution identities, and tool permissions.
硬體
Ultralytics 8.4.105 adds support for YOLO26 depth-estimation models on Hailo-8L and Qualcomm QNN, while improving deployment backends including CoreML, OpenVINO, and NCNN. The release also enables interrupted knowledge-distillation checkpoints to be validated and exported directly, although performance gains still need to be measured separately on the target hardware.
代理工具
Hallmark has seen rapid adoption on GitHub by packaging layout structures, typography, color, and anti-template checks into a Skill that Claude Code, Cursor, and Codex can load. Rather than training or fine-tuning models, it adds version-controlled design policies and self-review before and after generation.
最新模型
Anthropic has released Claude Opus 5 with a 1 million-token context window, up to 128,000 output tokens, and thinking enabled by default. The new model lets agents add or remove tools while preserving prompt caching, but disabling thinking is incompatible with high effort settings.
AI evaluation
Microsoft Research introduced EvolvingIntent, which transforms existing math, SQL, search, and programming tasks into multi-turn conversations in which requirements are progressively disclosed, revised, or redirected. Experiments show that models’ performance on single-turn benchmarks does not transfer reliably to scenarios where requirements continually change. The data-construction and evaluation code has been released publicly.
硬體
Microsoft Research has released VibeVoice-ASR-BitNet, applying INT8 and ternary-weight quantization to the speech encoder and language model, respectively, so multilingual ASR can run without a GPU. Official tests show the model is 1.6× to 2.3× faster than Whisper.cpp at a similar size, though it currently supports only offline batch transcription.