---
format: "aidr-story-markdown/v1"
id: "a4fc34b7908d0b559e32c5362da7e74259ad45d44b3cdaaac32d2e937bc20ee3"
canonical_url: "https://aidr.today/a4fc34b7?lang=en"
title: "Jeva: System One Models from First Principles"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-14T00:00:00.000Z"
category: "Research"
topics: ["inference"]
source_urls: ["https://huggingface.co/papers/2610.18773","https://huggingface.co/papers/2610.18615"]
summary: "Jeva is a decision-native architecture designed to address four structural requirements for system one models: null-awareness, permutation-invariance, option-scalability, and cacheability. It uses a shared encoder to independently represent requests and options, with a compact decision stack refining request state against immutable option memory. Symmetric aggregation produces permutation-equivariant probabilities, and a separate coverage estimate distinguishes answer absence from candidate competition."
---

# Jeva: System One Models from First Principles

> [Open the canonical story](<https://aidr.today/a4fc34b7?lang=en>)

**Published:** 2026-10-14T00:00:00.000Z
**Category:** Research
**Topics:** inference

## Summary

Jeva is a decision\-native architecture designed to address four structural requirements for system one models: null\-awareness, permutation\-invariance, option\-scalability, and cacheability\. It uses a shared encoder to independently represent requests and options, with a compact decision stack refining request state against immutable option memory\. Symmetric aggregation produces permutation\-equivariant probabilities, and a separate coverage estimate distinguishes answer absence from candidate competition\.

## Sources

- [Story source](<https://huggingface.co/papers/2610.18773>)
- [Story source](<https://huggingface.co/papers/2610.18615>)