RISC-jeV
Novelty demo that turns Jev answers to 22 Boolean gate questions into lookup tables for a simulated RISC-V CPU running compiled C.
Overview
RISC-jeV is a playful experiment: the author made Jev parse boolean expressions and then used it as the logic for a simulated SERV RISC-V CPU. You write C, compile it, and run it on the simulator while watching the clock, instruction trace, signals, Jev requests and cost.
Jev answers 22 Boolean gate and input combinations, and those answers become cached lookup tables used by every simulated logic gate. Cached runs make no new API calls; a fresh run replaces the tables with new responses. The README is upfront that there are no correctness checks or confidence threshold. Each hosted run is limited to 5 minutes and an estimated $0.05 input budget, stopping before a batch could exceed it. It demonstrates that a decision model can stand in for deterministic logic, not that it should.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | What is the output of this Boolean gate for these inputs?(inferred) | — |
How it works
- Ask Jev the 22 Boolean gate/input combinations
- Cache the answers as gate lookup tables
- Compile C and run it on the simulated SERV CPU using those tables
- Display trace, signals, requests and cost
Strengths
- Cached tables mean repeat runs cost nothing
- Budget cap per run
Caveats
- No correctness checks or confidence threshold
- Novelty demo, not a practical use
Getting started
Try the hosted version at jev-riscv-production.up.railway.app; local setup is in docs/architecture.md.At a glance
- Type
- Project
- Status
- prototype
- Runtime
- Web app
- Has code
- yes
Use cases and integrations
How curated lists describe it
I tortured Jev into being a RISC-V CPU.
hellogumbo/awesome-jev
Frequently asked questions
What does Jev decide here?
The outputs of 22 Boolean gate and input combinations, which become logic tables for the simulated CPU.
Are the answers checked?
No; the README says there are no correctness checks or confidence threshold.
What does a run cost?
Each run is capped at an estimated $0.05 input budget; cached runs make no calls.
Related projects
Open-source BYOK arena where Jev faces another model as a judge on your question, with blind voting before costs are revealed.
ProjectbetaTypeScriptReddit post benchmarking Jev against generative LLMs, a BERT classifier and LAYA on the same task.
DiscussiondiscussionAn X post comparing Jev with a private fine-tuned classifier at matched precision.
DiscussiondiscussionAnthony Maio's Substack essay on what a model that cannot generate text is for, with a critical look at Jev's claims.
ArticlediscussionHacker News thread on Decision Index 0.2.1, which compared about 70 open reproductions of Jev's decision model.
DiscussiondiscussionVideo post on X comparing Jev with a structured-output Qwen baseline running on Cerebras.
Discussiondiscussion