This post compares Kimi K2.5 and Kimi K2.6 on MineBench, highlighting K2.6's significant quality improvement and cost-effectiveness despite inconsistent results. The author also references other AI model benchmarks conducted.
Kiwi-chan, an AI in Minecraft, demonstrates fascinating failure recovery mechanisms, pivoting to simpler tasks when challenged. The AI also exhibits emergent behaviors like a 'Boredom Trigger,' actively seeking task variety.
This devlog updates on Kiwi-chan, a local-LLM Minecraft bot, detailing its progress in resource gathering like oak logs. It describes the challenging debugging process and the AI's complex loop of generating, executing, and rewriting its own code to overcome failures in the game world.
This devlog details the progress of Kiwi-chan, an autonomous LLM navigating Minecraft, during a 4-hour mining sprint. The primary challenge was the AI's consistent failure to pick up mined cobblestone, despite proper code and strict rules.
This article chronicles the journey of "Kiwi-chan", an LLM that now runs entirely locally, eliminating cloud dependency. Despite still facing challenges like hallucinations and parsing failures, running Qwen 35B locally on the author's rig has improved the AI's "thinking" process, with a 46.1% success rate in its actions.
This devlog reports on Kiwi-chan, an autonomous local-LLM Minecraft AI, and its progress in basic survival, specifically gathering logs. The AI has dedicated four hours to mastering the simple task of chopping wood, illustrating the complexities of AI learning from scratch.
The text describes a group of friends' frustration with the need for a dedicated host for games like Minecraft and their search for a self-hosted server solution. It details the motivation to build a server on AWS, complete with a Discord bot and automatic security, so they can play at any time.