OpenAI
AI Benchmarks Under Fire: Reward Hacks, IQ Scores, and Parameter Golf
A new site assigning IQ scores to 50+ AI models drew praise and sharp criticism this…
Tag: parameter golf
OpenAI
AI Benchmarks Under Fire: Hacking, IQ Scores, and Parameter
A new automated auditing tool called BenchJack found 219 reward-hacking exploits across 10 popular AI agent…
AI
AI IQ Site Ranks 50+ Models on Human Intelligence Scale
A new AI IQ platform ranking 50+ language models on human intelligence scales has sparked debate,…
AI
AI Architecture Advances: New Training Methods Cut GPU Waste
Enterprise GPU utilization averages just 5% despite $401 billion in AI spending, while new training techniques…