Cover image

Four Bits, Fewer Failure Modes

TL;DR for operators Post-training quantization methods that look similar on normal 4-bit benchmarks can separate sharply when the numerical problem becomes difficult. In one Mistral-7B v0.1 stress test, QuIP-family variants abort with NaNs and GPTQ reaches WikiText-2 perplexity of 379.73 ± 15.71. Kashin-DCT+H reaches 8.92 ± 0.01, close to the FP16 reference of 8.63. ...

October 20, 2026 · 8 min · Zelina

Free AI Inference Providers

A daily dashboard for monitoring free AI inference providers, with curated vendor boards and a machine-refreshable OpenRouter free-model roster.

April 1, 2026 · 1 min
Cover image

When Tokens Explode: The Hidden Geometry Behind Attention Sinks

Serving an LLM is usually discussed in pleasantly managerial language: latency, throughput, context windows, GPU memory, quantization, cache eviction. Nice clean nouns. Then the model ruins the spreadsheet by producing internal activations that are thousands of times larger than ordinary values, while some tokens quietly become attention magnets for reasons that are not exactly semantic. Very professional behavior from a trillion-dollar technology stack. ...

March 6, 2026 · 16 min · Zelina
Cover image

From Tokens to Teaspoons: What a Prompt Really Costs

TL;DR for operators A Google paper on Gemini Apps reports that the median text prompt in May 2025 consumed 0.24 Wh, generated 0.03 gCO2e, and consumed 0.26 mL of water under a comprehensive production-serving measurement boundary.1 That is small. Very small. Less “boil the kettle” and more “squint at a television for nine seconds.” ...

August 24, 2025 · 17 min · Zelina