Technology ❯ Artificial Intelligence

Performance Evaluation

Benchmarking Benchmarks Model Comparison Testing Methods Benchmark Testing Experimental Results State-of-the-Art Models Empirical Analysis Empirical Studies Reasoning Tasks IPC Domains Scalability Statistical Analysis Chatbots Benchmarking Platforms Generalization Error Rate Reduction Reasoning Efficiency Experimental Methods Accuracy Metrics Model Optimization Reasoning Tests Task Performance User Experience AI Testing Testing Metrics Conversational Skills Comparative Analysis

DeepSeek Releases V3.2 Models to Challenge GPT‑5 and Gemini 3 Pro

Independent validation of the company’s frontier‑level reasoning claims is still pending.

Anthropic’s Claude Sonnet 4.5 Often Spots Its Own Tests, Raising Evaluation Questions

New HANRAG Preprint Targets Multi‑hop Noise as RAG Guides Proliferate

Alibaba Debuts Trillion-Parameter Qwen-3-Max-Preview on Cloud and OpenRouter

Studies Unveil Scaling Laws and Edge Quantization for Mixture-of-Experts Language Models

Z.ai Unveils Open-Source GLM-4.5 Models That Undercut DeepSeek