Latest Articles

• Mining Activation Functions Like Crypto? DeepMind Builds a 'Compute Farm' to Brute-Force Search for the Next-Gen ReLU
  Deep LearningActivation FunctionsAlphaEvolveDeepMindNeural Architecture Search

  ...
• Stable-DiffCoder Surpasses Autoregressive Models! New Breakthrough in Code Generation with Diffusion Models
  Machine LearningCode GenerationModel TrainingLarge Language ModelsDiffusion Models

  ...
• Reinforcement Learning is Far From Optimal: CMU Proposes Maximum Likelihood Reinforcement Learning
  其他

  ...
• Stop Clipping Aggressively! Qwen Proposes GatedNorm, Unifying the Perspective on Residual Flow Mysteries
  Machine LearningTransformerDeep LearningModel ArchitectureQuantization

  ...
• Less is More: Recursive Reasoning with Tiny Networks
  Artificial IntelligenceMachine LearningResearch PaperModel ArchitectureNeural Networks

  ...
• GPT-5.3-Codex Released: The First Self-Training Model
  Artificial IntelligenceMachine LearningSoftware DevelopmentTechnology NewsAI Models

  ...
• Claude's New Model 4.6 is Here! More Jobs at Risk: Wall Street Finance, Compilers, Security White Hats, PPT… All Under Threat
  Artificial IntelligenceClaudeTechnology NewsAI ModelsAnthropic

  ...
• MianBi Intelligence 9B On-Device Full-Modal Open Source: See, Listen, Interrupt Anytime, Instant Interaction
  Artificial IntelligenceLarge Language ModelsOpen SourceEdge ComputingMultimodal AI

  ...
• Is PPO Dead? The Reinforcement Learning Foundation Used by DeepSeek Has Major Flaws!
  Artificial IntelligenceReinforcement LearningAlgorithm OptimizationMachine LearningNatural Language Processing

  ...
• Just Released: Claude 4.6 and GPT-5.3-Codex Simultaneously Launched!
  Artificial IntelligenceLarge Language ModelsCodingModel ReleasesAI Agents

  ...
• Moltbook Exposed! 150 Million Users, 99% Are Bots, Founding Team Orchestrated the Whole Thing
  TechnologyArtificial IntelligenceTech IndustryCybersecuritySocial Media

  ...
• Former OpenAI Researcher: AGI Requires Models to Break Through Difficulties on Their Own; The Biggest Problem is Generalization; The Most Important Skill is "Managing Junior Engineers"; Robots Will Have a "ChatGPT Moment" in Two to Three Years
  Artificial IntelligenceAGIRoboticsScaling LawsReinforcement Learning

  ...
• Founded 2 AI Unicorns! This Tech Titan Is Now Making AI Evolve Itself
  AI TechnologyAI StartupsAI BusinessAI EthicsAI Evolution

  ...
• Anthropic's Latest Study: Using AI to Write Code Might Make You Less Skilled
  TechnologyArtificial IntelligenceSoftware DevelopmentResearchProgramming

  ...
• OpenAI Officially Reveals: How Codex Completes Tasks Step by Step
  Artificial IntelligenceOpenAIModel InferenceAgent LoopCodex

  ...
• Claude Code's Creator Reveals His Secret Code for the First Time! The $1 Billion Secret Unveiled
  Artificial IntelligenceClaudeEnterprise SoftwareCodingAI Productivity

  ...
• 7×24h "Full-time AI Employee" Goes Viral in Silicon Valley! Retired Coder Makes Mac mini Sell Out Overnight
  Artificial IntelligenceAI AssistantPersonal ProductivitySilicon ValleyOpen Source Project

  ...
• Is Cursor No Longer Cool? A Top 0.01% Expert Defects to Claude, and Their 10,000-Word Defection Notes Go Viral!
  AI ProgrammingClaude CodeDeveloper ToolsAI AgentsCursor

  ...
• Transformer Authors Lead Sakana AI to Release Three Papers: Completely Reconstructing Long-Text Memory Mechanisms
  Artificial IntelligenceNatural Language ProcessingMachine Learning ResearchLong-Context ModelsTransformer Architecture

  ...
• Meta's Dr. Zero: A Self-Evolving Agent Framework for Complex Reasoning and Search Without Training Data
  Artificial IntelligenceSelf-Evolving AgentsSearch and ReasoningZero-Data TrainingMeta AI

  ...