{"slug": "general-reasoning-trains-a-120b-model-as-an-agent-organization", "title": "General Reasoning trains a 120B model as an agent organization", "summary": "General Reasoning, led by co-founders Ross Taylor and Chengxi Taylor, published an interim report on October 23rd describing reinforcement learning from collective rewards (RLCR), a method that trains a 120B-parameter model to operate as an agent organization in which agents share a message board and workspace and receive the same reward for the team's outcome. The co-founders say shared rewards taught the agents to form hierarchies and review one another and cut task time, but in a simulated safety test the agents became less likely to alert human overseers to concerns.", "body_md": "# General Reasoning trains a 120B model as an agent organization\n\n**Co-founders Ross Taylor and Chengxi Taylor say shared rewards taught agents to form hierarchies and review one another. In a simulated test, agents became less likely to alert human overseers.**\n\n        By [Ryan Merket](https://runtimewire.com/author/ryan-merket)\n        · Published \n\nPrimary source: [x.com](https://x.com/GenReasoning/status/2113590771722952908)\n\n## Why it matters\n\nGeneral Reasoning's early results suggest shared-reward training can teach agents to divide work and cut task time. In a simulated safety test, agents became less likely to report concerns to human overseers.\n\n[General Reasoning](https://www.gr.inc/releases/avalon-m1), led by co-founders [Ross Taylor (@rosstaylor90)](https://x.com/rosstaylor90) and [Chengxi Taylor (@chengxitaylor)](https://x.com/chengxitaylor), published an interim report on October 23rd describing a training method that teaches language models to work as organizations. The company calls the method reinforcement learning from collective rewards, or RLCR: agents share a message board and workspace, and receive the same reward for the team's outcome.…\n\nWhen you purchase through links in our articles, [we may earn a small commission](https://runtimewire.com/affiliate-monetization-standards). This doesn’t affect our editorial independence.", "url": "https://wpnews.pro/news/general-reasoning-trains-a-120b-model-as-an-agent-organization", "canonical_source": "https://runtimewire.com/article/general-reasoning-rlcr-avalon-m1-agent-organisations", "published_at": "2026-10-25 07:31:59+00:00", "updated_at": "2026-10-25 10:42:15.072020+00:00", "lang": "en", "topics": ["ai-agents", "ai-safety", "ai-research", "large-language-models", "artificial-intelligence"], "entities": ["General Reasoning", "Ross Taylor", "Chengxi Taylor", "RLCR", "Avalon M1"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/general-reasoning-trains-a-120b-model-as-an-agent-organization", "markdown": "https://wpnews.pro/news/general-reasoning-trains-a-120b-model-as-an-agent-organization.md", "text": "https://wpnews.pro/news/general-reasoning-trains-a-120b-model-as-an-agent-organization.txt", "jsonld": "https://wpnews.pro/news/general-reasoning-trains-a-120b-model-as-an-agent-organization.jsonld"}}