Verantyx

Symbolic Reasoning. Zero Neural Networks.

No LLMs. No pretrained models. No neural networks.

Every solution is interpretable and verifiable.

Pure symbolic program synthesis

Quick Install
git clone https://github.com/Ag3497120/verantyx-v6 && cd verantyx-v6 && pip install -r requirements.txt
ARC-AGI-2
0.0%
207/1000 tasks
HLE
0.0%
Hard Logic Evaluation
Zero neural networks. Every solution is a verifiable program.
Phase 1: Cross DSL

How It Works

Zero LLMs, zero neural networks — every solution is a verifiable program

Input Grid

Piece Generation

Cross DSL

Verification

Output

Cross DSL Approach

Verantyx uses neighborhood-rule transformations to systematically explore the space of possible programs. Each transformation is a composable, verifiable operation — no probabilistic guessing, no gradient descent.

Verifiable by Design

Every generated program can be inspected, understood, and verified. No black boxes. No unexplainable behavior. Just pure, deterministic symbolic reasoning.

7-Phase Symbolic Pipeline

1

Cross DSL

Neighborhood Rules
57% of solutions
2

Standalone Primitives

flip, rotate, crop
Pure transformations
3

Stamp

Pattern Fill
Template matching
4

Composite Chains

Multi-step pipelines
Sequential operations
5

Iterative Cross

Residual learning
Recursive refinement
6

Puzzle Reasoning Language

High-level abstractions
Symbolic logic
7

ProgramTree Synthesis

CEGIS
Counter-example guided
Each phase is a nested cross structure — symbolic reasoning all the way down

Evolution

From 11.3% to 20.7% — continuous improvement through symbolic refinement

0%5%10%15%20%25%11.3%v1913.8%v2515.5%v3217.2%v4018.9%v4720.7%v53

Each version introduces new symbolic transformations and reasoning strategies

304
files
100K
lines
0.48s
per task
0
dependencies

Pure symbolic reasoning. No neural networks.