Agents & Tool Use · 2025
Kimi-Researcher: End-to-End RL Training for Emerging Agentic Capabilities
Trains an autonomous research agent end-to-end by agentic RL to plan, search, and use tools over long horizons, letting strategies emerge from outcome rewards rather than hand-designed workflows.
Editorial record
Plain-language summary
Trains an autonomous research agent end-to-end by agentic RL to plan, search, and use tools over long horizons, letting strategies emerge from outcome rewards rather than hand-designed workflows.
Knowledge graph
Relationships
Descendants
Conceptual ancestorEvidence: Strongly supported
ReAct: Synergizing Reasoning and Acting in LMs
Builds on the agents lineage in the archive
freshness sweep 2026
Source record
Provenance
- Record ID
- P-649
- Record created
- 2026-07-13
- Last reviewed
- 2026-07-14
- Record version
- 2
Citation caveat: Citation metadata is approximate and marked unverified in the source dataset.