--- title: "#model-evaluation articles | Hashnode" description: "Explore 78 articles tagged #model-evaluation on Hashnode. Tutorials, guides, and how-tos from the community." url: "https://hashnode.com/tag/model-evaluation" --- Tag # #model-evaluation 78 posts · 0 followers [Vashni Roy](https://hashnode.com/@vashniroy) · [1d ago](https://hashnode.com/posts/why-0-5-isn-t-always-the-right-classification-threshold/6af2ffa99538fc7c0f2ddfef) ### [Why 0.5 Isn't Always the Right Classification Threshold](https://vashniroy.hashnode.dev/why-0-5-isn-t-always-the-right-classification-threshold?utm_source=hashnode&utm_medium=feed) I changed the threshold of the same churn model from 0.3 to 0.7. The model stayed the same, but its predictions changed quite a lot. 3 min read · [vashniroy.hashnode.dev](https://vashniroy.hashnode.dev/why-0-5-isn-t-always-the-right-classification-threshold?utm_source=hashnode&utm_medium=feed) [Future Tech Career Hub](https://hashnode.com/@futuretechcareerhub) · [Oct 19](https://hashnode.com/posts/data-science-models-fail-in-production/6ad5f53e9c05e76fc0086220) ### [A Model That Works in Testing Can Still Fail in Production: Lessons for Data Scientists](https://futuretechcareerhub.hashnode.dev/data-science-models-fail-in-production?utm_source=hashnode&utm_medium=feed) A Model That Works in Testing Can Still Fail in Production: Lessons for Data Scientists 10 min read · [futuretechcareerhub.hashnode.dev](https://futuretechcareerhub.hashnode.dev/data-science-models-fail-in-production?utm_source=hashnode&utm_medium=feed) [Vashni Roy](https://hashnode.com/@vashniroy) · [Oct 14](https://hashnode.com/posts/does-a-more-complex-ml-model-actually-give-better-results/6acf5b94f069b0121b58b0e5) ### [Does a More Complex ML Model Actually Give Better Results?](https://vashniroy.hashnode.dev/does-a-more-complex-ml-model-actually-give-better-results?utm_source=hashnode&utm_medium=feed) I compared Logistic Regression and XGBoost on the same churn dataset. The simpler model performed better, even across 5-fold cross-validation. 3 min read · [vashniroy.hashnode.dev](https://vashniroy.hashnode.dev/does-a-more-complex-ml-model-actually-give-better-results?utm_source=hashnode&utm_medium=feed) [Gary Stafford](https://hashnode.com/@garystafford) · [Oct 6](https://hashnode.com/posts/how-to-set-up-evals-for-a-complex-multi-agent-system/6ac46769848e3306fd9c5c03) ### [How to Set Up Evals for a Complex Multi-Agent System](https://blog.latentthoughts.com/how-to-set-up-evals-for-a-complex-multi-agent-system?utm_source=hashnode&utm_medium=feed) Using Codex and Promptfoo to define success, diagnose failures, and choose models 29 min read · [blog.latentthoughts.com](https://blog.latentthoughts.com/how-to-set-up-evals-for-a-complex-multi-agent-system?utm_source=hashnode&utm_medium=feed) [5](https://hashnode.com/posts/how-to-set-up-evals-for-a-complex-multi-agent-system/6ac46769848e3306fd9c5c03) [Gary Stafford](https://hashnode.com/@garystafford) · [Oct 2](https://hashnode.com/posts/choosing-the-right-openai-model-and-reasoning-level/6abf0d0682c3fdfa1bc9ab22) ### [Choosing the Right OpenAI Model and Reasoning Level](https://blog.latentthoughts.com/choosing-the-right-openai-model-and-reasoning-level?utm_source=hashnode&utm_medium=feed) An evaluation guide to accuracy, latency, and cost across complex workloads. 20 min read · [blog.latentthoughts.com](https://blog.latentthoughts.com/choosing-the-right-openai-model-and-reasoning-level?utm_source=hashnode&utm_medium=feed) [6](https://hashnode.com/posts/choosing-the-right-openai-model-and-reasoning-level/6abf0d0682c3fdfa1bc9ab22) [Muhammad Ariel Shakaramiro](https://hashnode.com/@arielshakaramiro) · [Sep 15](https://hashnode.com/posts/single-object-detection-why-boxes-are-off/6aa972d1a9dd656f74e1fd22) ### [Single Object Detection: Why the Bounding Box Wasn't Right](https://shaka-ai.hashnode.dev/single-object-detection-why-boxes-are-off?utm_source=hashnode&utm_medium=feed) Three iterations of an object detector and one unexpected finding: the model may be outputting a near-constant guess, not precise object location. 11 min read · [shaka-ai.hashnode.dev](https://shaka-ai.hashnode.dev/single-object-detection-why-boxes-are-off?utm_source=hashnode&utm_medium=feed) [Munukutla Sri Shanmukha Aditya](https://hashnode.com/@m-s-shanmukha-aditya) · [Sep 4](https://hashnode.com/posts/why-accuracy-can-be-misleading-in-failure-detection-systems/6a9b006c56ada066e5721232) ### [Why Accuracy Can Be Misleading in Failure Detection Systems](https://shanmukha-aditya.hashnode.dev/why-accuracy-can-be-misleading-in-failure-detection-systems?utm_source=hashnode&utm_medium=feed) Why accuracy can mislead in failure detection systems, and why precision and recall matter when working with imbalanced data. 4 min read · [shanmukha-aditya.hashnode.dev](https://shanmukha-aditya.hashnode.dev/why-accuracy-can-be-misleading-in-failure-detection-systems?utm_source=hashnode&utm_medium=feed) [1](https://hashnode.com/posts/why-accuracy-can-be-misleading-in-failure-detection-systems/6a9b006c56ada066e5721232) [Quinn](https://hashnode.com/@quinn-yang) · [Jul 31](https://hashnode.com/posts/a-repeatable-way-to-compare-ai-video-models-without-cherry-picking/6a6cc41b908f19383269f3d2) ### [A Repeatable Way to Compare AI Video Models Without Cherry-Picking](https://mict-engineering-notes.hashnode.dev/a-repeatable-way-to-compare-ai-video-models-without-cherry-picking?utm_source=hashnode&utm_medium=feed) A small test matrix, a precommitted rubric, and a cost-aware decision rule 9 min read · [mict-engineering-notes.hashnode.dev](https://mict-engineering-notes.hashnode.dev/a-repeatable-way-to-compare-ai-video-models-without-cherry-picking?utm_source=hashnode&utm_medium=feed) [Sunny Gupta](https://hashnode.com/@sunnykgupta) · [Jul 28](https://hashnode.com/posts/seven-models-in-seven-days-how-to-pick-without-whiplash/6a6912fc85793bfbc8868033) ### [Seven Models in Seven Days: How to Pick Without Whiplash](https://icodestartups.com/seven-models-in-seven-days-how-to-pick-without-whiplash?utm_source=hashnode&utm_medium=feed) Seven AI models shipped in seven days this month. Then Anthropic closed the week with Opus 5, near-frontier intelligence at half the flagship price. 7 min read · [icodestartups.com](https://icodestartups.com/seven-models-in-seven-days-how-to-pick-without-whiplash?utm_source=hashnode&utm_medium=feed) [1](https://hashnode.com/posts/seven-models-in-seven-days-how-to-pick-without-whiplash/6a6912fc85793bfbc8868033) [Manu Shukla](https://hashnode.com/@mr-manushukla) · [Jul 29](https://hashnode.com/posts/4-metrics-that-beat-benchmarks-the-enterprise-ai-model-scorecard-2026/6a69dda9add2f4bd69d2670a) ### [4 metrics that beat benchmarks: the enterprise AI model scorecard (2026)](https://ecorpit.hashnode.dev/4-metrics-that-beat-benchmarks-the-enterprise-ai-model-scorecard-2026?utm_source=hashnode&utm_medium=feed) 4 metrics that beat benchmarks: the enterprise AI model scorecard (2026) Summary. In 2026 the benchmark scoreboard stopped telling you which model to ship. MMLU has saturated above 90%, frontier model 13 min read · [ecorpit.hashnode.dev](https://ecorpit.hashnode.dev/4-metrics-that-beat-benchmarks-the-enterprise-ai-model-scorecard-2026?utm_source=hashnode&utm_medium=feed)