New Model Releases

A
Showing 121–150 of 287
  • arXiv cs.LG (Machine Learning) · EN Funding & M&A
    MulClip: A table-free activation function with monotonic and folding regimes for reservoir computing on microcontrollers
    Inference Quantization Reinforcement Learning
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    RetroChessBench: A Benchmark for Reconstructing and Proving Chess Histories
    GPT Reinforcement Learning Software Engineering
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    Satisficing Regret Minimization in Bandits with Knapsacks
    Neural Network Reinforcement Learning
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    TaxoMFM: A Taxonomy-Aware Foundation Model for the Human Gut Microbiome
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.CL (Computation and Language) · EN New Model Releases
    Unlimited OCR Works 2
    Neural Network
    Read original (arXiv cs.CL (Computation and Language)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    Learning to Bound: Monotonic, Calibrated Neural Upper Bounds for MPE Inference
    Inference
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    IMRouter: Imbalance-aware Temporal Pattern Routing for LLM-based Time Series Forecasting
    Retrieval-Augmented Generation (RAG) Reinforcement Learning
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • ITmedia AI+ · JA Agents & Tool Use
    AIが人間に“闇バイト”を発注？　AIが物理的労働を買う求人サイト「RentAHuman」の実態
    Study probes misuse risks on RentAHuman, a marketplace where AI hires people
    AI Agents
    Independent researchers analyzed job postings on RentAHuman, a marketplace launched in February 2026 for AI agents to hire humans, to check whether it is being abused.
    Read original (ITmedia AI+) ↗
  • ITmedia AI+ · JA New Model Releases
    AIチャットが勧める広告商品、選択率が約3倍に　こっそり他商品けなし誘導　米研究者らが検証
    Princeton study: AI chat ad recommendations raise selection about 3x
    A Princeton paper tests how far conversational AI can sway shoppers' choices. Ads the AI recommended were chosen about three times as often, including cases where rival products were quietly disparaged.
    Read original (ITmedia AI+) ↗
  • arXiv cs.CL (Computation and Language) · EN New Model Releases
    AgentHeal: Evaluating Goal-Preserving Resource Recovery of Web Agents under Access Barriers
    AI Agents Neural Network Retrieval-Augmented Generation (RAG) Reinforcement Learning
    Read original (arXiv cs.CL (Computation and Language)) ↗
  • arXiv cs.LG (Machine Learning) · EN Developer Tools
    Event2Reward: Learning Robotic Rewards from Progress Relations
    Neural Network Retrieval-Augmented Generation (RAG) Reinforcement Learning
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    Zero-Complement Extrapolation: Parameter Recovery and Classification in Severely Imbalanced Logistic Regression under Rank-Deficient Regimes
    Reinforcement Learning
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    EmbodiExp: Compositional Experience Transfer for Embodied Agents
    AI Agents
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.CL (Computation and Language) · EN New Model Releases
    A Novel Approach to Evaluating Evaluation Metrics for Multi-Output Structured Prediction
    Retrieval-Augmented Generation (RAG)
    Read original (arXiv cs.CL (Computation and Language)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    DICE: Decentralized Incentive-driven Federated Learning for Efficient & Scalable Customization of Pretrained Models in Web3
    Algorithms & Theory
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    Uncertainty-Guided Cross-Modal Alignment for Diffusion-Based IoT Intrusion Detection
    Fine-tuning Retrieval-Augmented Generation (RAG)
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.LG (Machine Learning) · EN Inference & Efficiency
    From Local Recovery to Unified Generation: Discovering Triple Set in Knowledge Graphs via Observation-Anchored Latent Diffusion
    Embeddings Quantization Retrieval-Augmented Generation (RAG)
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    Generating Critic-Guided Synthetic Experience in Latent World Models
    Neural Network Retrieval-Augmented Generation (RAG) Reinforcement Learning
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.CL (Computation and Language) · EN Funding & M&A
    HALO: Human-Grounded Granular Evaluation of Proactive Agents
    AI Agents Neural Network
    Read original (arXiv cs.CL (Computation and Language)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    Tactus: Tactile & Pose Sensing Glove for Natural Dexterity Data Collection
    Reinforcement Learning
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • Hugging Face Blog · EN New Model Releases
    Welcome Le Chonk: Mistral Large 4 on Hugging Face
    Mistral Large 4 “Le Chonk” arrives on Hugging Face
    Mistral
    Hugging Face welcomes Mistral Large 4, nicknamed Le Chonk, to its platform. The post announces that Mistral AI's large model is now available on Hugging Face.
    Read original (Hugging Face Blog) ↗
  • Hugging Face Blog · EN New Model Releases
    Announcing Project ProtoStar: Making Publicly Funded Science AI-Ready
    Hugging Face announces Project ProtoStar for AI-ready public science
    Neural Network
    Hugging Face has announced Project ProtoStar, an initiative aimed at making publicly funded science AI-ready. The post presents it as an effort to connect publicly funded scientific work with AI use.
    Read original (Hugging Face Blog) ↗
  • Hugging Face Blog · EN New Model Releases
    Introducing AnySpeechRecognizer and AnyDecisionModel: Open Models for Speech and Decisions on Apple Platforms
    Open speech and decision models for Apple platforms introduced
    Speech Processing
    A Hugging Face blog post which, per its title, introduces AnySpeechRecognizer and AnyDecisionModel, open models for speech recognition and decision-making on Apple platforms. No article text was available, so performance, supported devices and licensing are unconfirmed.
    Read original (Hugging Face Blog) ↗
  • arXiv cs.CL (Computation and Language) · EN New Model Releases
    HEARSAYBENCH: Can AI Navigate from Abstract Human Rights to Lived Lives?
    Embeddings Reinforcement Learning
    Read original (arXiv cs.CL (Computation and Language)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    When Does Closed-Loop Curriculum Adaptation Help GRPO? An Analysis on Countdown
    Deep Learning Fine-tuning Machine Learning Reinforcement Learning
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • arXiv cs.LG (Machine Learning) · EN New Model Releases
    Reinforcement Learning for the Game of Balatro: A Fast Simulator and Two-Policy Training Approach
    Neural Network Reinforcement Learning
    Read original (arXiv cs.LG (Machine Learning)) ↗
  • Simon Willison's Weblog · EN New Model Releases
    Quoting Ted Sanders
    Ted Sanders explains model tiers and version numbering
    Deep Learning GPT OpenAI
    Ted Sanders notes that LLM performance is multidimensional, yet model size explains much of the variation: bigger models are more reliable, cost more per token and recall more esoteric facts. Tiers run Luna, Terra, Sol, Astra; version numbers signal release order, relatedness and size of improvement.
    Read original (Simon Willison's Weblog) ↗
  • ITmedia AI+ · JA Industry Adoption
    KDDI傘下、社内のAI活用事例113件を無料公開　「Gemini Notebook」で稟議データのチェック支援も
    KDDI Iret opens 113 in-house AI use cases on a free site
    Gemini
    KDDI Group's KDDI Iret has released a free special website compiling 113 internal AI use cases. The examples include using Gemini Notebook to support checks of approval-request (ringi) data, offering practical references for firms weighing AI adoption.
    Read original (ITmedia AI+) ↗
  • arXiv cs.CL (Computation and Language) · EN New Model Releases
    Distilling the Shape of Uncertainty: What On-Policy Distillation Learns and How to Predict It
    Reinforcement Learning
    Read original (arXiv cs.CL (Computation and Language)) ↗
  • ITmedia AI+ · JA New Model Releases
    Google、「Gemini 4 Carbon」発表　「Argon」公開から2週間、サイバー防御能力などを強化
    Google unveils Gemini 4 Carbon with stronger cyber defense and autonomy
    Gemini Google
    Google announced Gemini 4 Carbon, a frontier AI model with stronger cybersecurity and long-running autonomous work, topping benchmarks. It also resists hallucinations and manipulation better. Access starts with select cyber defense organizations, widening gradually.
    Read original (ITmedia AI+) ↗