# Position Paper Argues AI Reasoning Evaluations Lack a Working Definition of Reasoning

The authors propose operational definitions that treat valid reasoning as a learnable, rule-based process, along with a reporting checklist, arguing that current benchmarks cannot show they actually measure reasoning rather than something else.

- Published: 2026-08-15T06:14:29.474Z
- Canonical: https://polylog.news/ai/2026-08-15/position-paper-argues-ai-reasoning-evaluations-lack-a-workin
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [arXiv cs.AI](https://arxiv.org/abs/2608.12325</source-url-placeholder)

A position paper on arXiv, Reasoning is a Learnable Rule-Based Process, targets a problem that underlies every reasoning benchmark result published this year. The authors argue that the AI and generative AI communities use the word reasonin…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-08-15/position-paper-argues-ai-reasoning-evaluations-lack-a-workin (subscription information: https://polylog.news/pricing).