NLP Lineage

Research thread · 2021 to 2026

Agents

36 milestones sit on this thread. For 6 of them it is the main thread. The others also belong to another thread, which is named under them.

2021

  • CopilotGitHub CopilotProduct
  • WebGPTWebGPT: Browser-assisted question-answering with human feedbackMain thread: Representation and retrievalResearch paper

2022

  • ReActReAct: Synergizing Reasoning and Acting in Language ModelsMain thread: Representation and retrievalTechnique

2023

  • ToolformerToolformer: Language Models Can Teach Themselves to Use ToolsMain thread: Representation and retrievalResearch paper
  • ReflexionReflexion: Language Agents with Verbal Reinforcement LearningMain thread: ReasoningResearch paper
  • AutoGPTAutoGPTInfrastructure
  • Tree of thoughtsTree of Thoughts: Deliberate Problem Solving with Large Language ModelsMain thread: ReasoningTechnique
  • Function callingFunction calling in the OpenAI APIMain thread: Representation and retrievalProduct
  • SWE-benchSWE-bench: Can Language Models Resolve Real-World GitHub Issues?Main thread: ReasoningDataset

2024

  • DevinDevinProduct
  • SWE-agentSWE-agent: Agent-Computer Interfaces Enable Automated Software EngineeringMain thread: Representation and retrievalResearch paper
  • Computer useClaude computer useMain thread: MultimodalityProduct
  • MCPModel Context ProtocolMain thread: Representation and retrievalInfrastructure
  • Gemini 2.0Gemini 2.0Main thread: MultimodalityModel

2025

  • Deep researchDeep researchMain thread: Representation and retrievalProduct
  • Claude CodeClaude CodeMain thread: Representation and retrievalProduct
  • o3 releasedOpenAI o3Main thread: ReasoningModel
  • Claude 4Claude 4 (Opus 4 and Sonnet 4)Main thread: ReasoningModel
  • Kimi K2Kimi K2Main thread: Language modellingModel
  • GPT-5GPT-5Main thread: ReasoningModel
  • GLM-4.5GLM-4.5Main thread: ReasoningModel
  • Claude Sonnet 4.5Claude Sonnet 4.5Main thread: ReasoningModel
  • Gemini 3Gemini 3Main thread: ReasoningModel
  • Claude Opus 4.5Claude Opus 4.5Main thread: ReasoningModel

2026