// guides · how to
How to
This topic holds the TrueIQ how to guides on AI benchmarks and evaluation: practical walkthroughs that start from a clear goal and end with something working. Each one lists what you need before you begin, gives the steps in order and links to the official documentation it was checked against.
There are 15 guides here, published from Oct 06, 2026 to Nov 12, 2026. The tools they cover most often are DeepEval, GPTZero and Langfuse.
How to choose: pick the guide whose title matches the task, not just the tool, because many tools appear in several guides. Check the last checked date at the top, since menus and setting names change between versions. If a step no longer matches what you see, the contact page is the quickest way to tell us.
10 min readListen
How to Use GPTZero in Google Docs (2026)
How to use GPTZero in Google Docs: install the free extension, scan from the GPTZero menu, highlight AI sentences and open a shareable Writing Report.
9 min readListen
How to Spot AI Scam Ads on Facebook and YouTube
How to spot AI scam ads that use deepfake celebrities on Facebook, Instagram and YouTube: the red flags, a quick check, how to report them and what detectors miss.
9 min readListen
How to Evaluate LLMs Before Production
How to evaluate LLMs before production: set the product decision, build a production like offline suite, wire CI gates, then open shadow or A/B traffic.
9 min readListen
How to Use the PyRIT Tool: A Tested Walkthrough
How to use the PyRIT tool from Microsoft: install it, configure a target, run your first attack with converters and a scorer, then try the scanner and Crescendo.
10 min readListen
How to Use Promptfoo: A Tested Walkthrough
How to use Promptfoo, step by step: install it, write a promptfooconfig.yaml, add assertions, compare two versions, read results and gate CI on the pass rate.
· updated
10 min readListen
How to Check If a Video Is AI Generated (2026)
How to check if a video is AI generated in 2026: platform labels, Content Credentials, SynthID Detector, classifiers and the source checks that still matter.
· updated
10 min readListen
How to Use LangSmith with LangGraph (2026)
How to use LangSmith with LangGraph: set four environment variables, read node traces and threads, then score your graph and single nodes on a dataset.
9 min readListen
How to Use the MMLU Benchmark, Step by Step
How to use MMLU benchmark runs: pick the right variant, run it with lm-evaluation-harness or Inspect, smoke test the format, and read the score honestly.
10 min readListen
How to Detect AI Writing: A Step by Step Guide
How to detect AI writing: read for the patterns, check facts and sources, run two detectors, ask for drafts, then talk to the writer. A careful method, free.
9 min readListen
How to Take a Coding Test: Step by Step
How to take a coding test: check your setup, read every task, submit a correct solution early, then test edge cases and speed, using HackerRank and Codility rules.
9 min readListen
How to Evaluate LLM Model Performance
How to evaluate LLM model performance before you pick one: build a task test set, fix settings, add error bars, test prompt sensitivity, then weigh cost.
· updated
7 min readListen
How to Use Originality.ai for Free
How to use Originality.ai for free in 2026: 100 free credits per account (about 10,000 words), how credits burn by model, privacy notes and how to read AI scores.
· updated
8 min readListen
How to Use Pangram on Substack
How to use Pangram on Substack: Scan for AI text on web and iOS, creator controls, July 2026 limits, and how to read the human vs AI assisted estimate.
· updated
8 min readListen
SynthID Detector: How to Check for AI Watermarks
Google's SynthID Detector is now open to everyone. How to check images, video and audio for its AI watermark, what results mean, the limits, and alternatives.
· updated
9 min readListen
How to Evaluate LLM Applications (2026 Guide)
How to evaluate LLM applications step by step: error analysis, code checks, calibrated LLM judges, CI gates and production sampling, with tools for each stage.
· updated














