Gemini 3.6 Flash Deprecates Sampling Controls: Update Grounded-Answer Test Harnesses
Migrate Gemini test harnesses after temperature, top_p, and top_k deprecation while preserving model, grounding, citation, latency, and usage evidence.
Tagged collection
Latest articles tagged “LLM Evaluation”
Open an article for the complete method, evidence, limitations, and a focused reader discussion.
-
-
GPT-5.6 Fast Mode Above 272K Tokens: A Latency-Cost Test Protocol
A preregistered protocol for comparing GPT-5.6 Fast and standard service above 272K input tokens across latency, cost, errors, and task quality.
-
chat-latest Is a Moving Target: Freeze Model IDs for Reproducible Tests
OpenAI says chat-latest is regularly updated. Use a fixed model for regression tests and a complete run manifest when comparing current ChatGPT-aligned behavior.
-
How to Read Brave’s 1,500-Query AI Search Benchmark
Audit Brave’s 1,500-query AI answer comparison by examining query sampling, interface parity, LLM judges, pairwise order, ownership, and replication.
