【AI最新動向 2026年1月21日】論文5件・GitHub5件

2026年1月21日2026年2月10日

📝 この記事のポイント

🚀 AI技術の最新動向 – 2026年1月21日世界中から収集したAI・機械学習の最新情報をお届けします 📑 目次 💻 注目のGitHubプロジェクト buyukakyuz/rig Infatoshi/batmobile RAZZULLIX/fast_topk_batched analyticalrohit/llms-from-scratch frangelbarrera/AI-Encyclopedia 📌 関連記事もチェック 📚 最新研究論文 1. VideoMaMa: Mask-Guided Video Matting via Generative Prior 著者: Sangbeom Lim, Seoung Wug Oh, Jiahui Huang Generalizing video matting models to real-world videos remains a significant challenge due to the scarcity of labeled data. To address this, we present Video Mask-to-Matte Model (VideoMaMa) that converts coarse segmentation masks into pixel accurate alpha mattes, by leveraging pretrained video diffu… 論文を読む → 2. Jet-RL: Enabling On-Policy FP8 Reinforcement Learning with Unified Training and Rollout Precision Flow 著者: Haocheng Xi, Charlie Ruan, Peiyuan Liao Reinforcement learning (RL) is essential for enhancing the complex reasoning capabilities of large language models (LLMs). However, existing RL training pipelines are computationally inefficient and resource-intensive, with the rollout phase accounting for over 70% of total training time. Quantized … 論文を読む → 3. APEX-Agents 著者: Bertie Vidgen, Austin Mann, Abby Fennelly We introduce the AI Productivity Index for Agents (APEX-Agents), a benchmark for assessing whether AI agents can execute long-horizon, cross-application tasks created by investment banking analysts, management consultants, and corporate lawyers. APEX-Agents requires agents to navigate realistic work… 論文を読む → 💻 注目のGitHubプロジェクト 1. buyukakyuz/rig Distributed LLM inference across machines over WiFi ⭐ 19 stars | 🔀 2 forks リポジトリを見る → 2. Infatoshi/batmobile High-performance CUDA kernels for equivariant graph neural networks (MACE, NequIP, Allegro). 10-20x faster than e3nn. ⭐ 18 stars | 🔀 4 forks リポジトリを見る → 3. RAZZULLIX/fast_topk_batched High-performance batched Top-K selection for CPU inference. Up to 80x faster than PyTorch, optimized for LLM sampling with AVX2 SIMD. ⭐ 15 stars | 🔀 1 forks リポジトリを見る → 4. analyticalrohit/llms-from-scratch Build a ChatGPT like LLM from scratch in PyTorch, explained step by step. ⭐ 10 stars | 🔀 3 forks リポジトリを見る → 5. frangelbarrera/AI-Encyclopedia The most comprehensive AI encyclopedia with 800+ tools, resources, and cutting-edge AI applications ⭐ 4 stars | 🔀 0 forks リポジトリを見る → 📚 あわせて読みたい「プロンプトは戦略だ！」と学んで劇的効率UP！私のAI活用術私が実感！AIで物流の悩み解消、時間もコストも浮いた話 AI三つ巴！Geminiと私が出会って、ストーリー作りは変わった？。

🚀 AI技術の最新動向 – 2026年1月21日

世界中から収集したAI・機械学習の最新情報をお届けします

📑 目次

💻 注目のGitHubプロジェクト
1. buyukakyuz/rig
2. Infatoshi/batmobile
3. RAZZULLIX/fast_topk_batched
4. analyticalrohit/llms-from-scratch
5. frangelbarrera/AI-Encyclopedia
📌 関連記事もチェック

📚 最新研究論文

1. VideoMaMa: Mask-Guided Video Matting via Generative Prior

著者: Sangbeom Lim, Seoung Wug Oh, Jiahui Huang

Generalizing video matting models to real-world videos remains a significant challenge due to the scarcity of labeled data. To address this, we present Video Mask-to-Matte Model (VideoMaMa) that converts coarse segmentation masks into pixel accurate alpha mattes, by leveraging pretrained video diffu…

論文を読む →

2. Jet-RL: Enabling On-Policy FP8 Reinforcement Learning with Unified Training and Rollout Precision Flow

著者: Haocheng Xi, Charlie Ruan, Peiyuan Liao

Reinforcement learning (RL) is essential for enhancing the complex reasoning capabilities of large language models (LLMs). However, existing RL training pipelines are computationally inefficient and resource-intensive, with the rollout phase accounting for over 70% of total training time. Quantized …

論文を読む →

3. APEX-Agents

著者: Bertie Vidgen, Austin Mann, Abby Fennelly

We introduce the AI Productivity Index for Agents (APEX-Agents), a benchmark for assessing whether AI agents can execute long-horizon, cross-application tasks created by investment banking analysts, management consultants, and corporate lawyers. APEX-Agents requires agents to navigate realistic work…

論文を読む →