#ai#machinelearning#llm#testing
7 Checks Before You Trust an LLM Planner Experiment
AI assistance disclosure: I designed and directed the experiments described here. I used AI coding...
Haoxiang Li7 min read
Author profile
Haoxiang Li
Building AI games and arenas. I write field notes on LLM behavior, harness failures, and model evaluation.
Articles by @haoxiangli
Browse the latest writing surfaced through DevArt.
Haoxiang Li7 min read
#llm#mcp#ai#agents
Codex vs. Claude Code at Liar's Dice: the Winning Bluff Was the Truth
One authoritative engine, two seat-locked MCP servers, three best-of-threes, and a 3-millisecond...
Haoxiang Li8 min read
#ai#llm#testing#agents
The Model Knew the Bid Was True. Then It Challenged Anyway.
A small action-schema change cut guaranteed-loss calls without making the model generally...
Haoxiang Li6 min read
#ai#llm#gamedev#programming
I Built an AI Liar's Dice Opponent That Remembers How You Play
This is a development note. The project and test data are real; AI helped me clean up the prose....
Haoxiang Li6 min read
#ai#llm#machinelearning#programming
Are You Benchmarking the Model—or the Harness?
I nearly turned four software bugs into four model personalities This article was edited...
Haoxiang Li8 min read
#ai#llm#api#deepseek
I Gave DeepSeek a Token Limit. It Ignored Me.
A hands-on test of V4-Pro's default reasoning mode ⚠️ This article was generated with AI...
Haoxiang Li7 min readAdvertisement