Skip to content

netover/wave-field-llm

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 

Repository files navigation

Wave Field LLM v6

A novel attention mechanism for language models that achieves O(N log N) complexity, replacing the standard O(N²) dot-product attention.

Tokens interact through wave propagation on a continuous field rather than direct pairwise comparison — enabling efficient scaling to long contexts that are impossible with standard transformers.

Key Results

Metric Value
DCLM CORE (130M model) 46.8% (GPT-2 target: 26.5%)
Throughput at 32K context 21.8x faster than standard
Memory at 32K context 5.3x less than standard
128K context Runs (standard OOMs)
Model size 505 MB (runs on laptop)

See BENCHMARKS.md for full results.

How It Works

  1. Scatter — tokens deposit information onto a continuous field
  2. Convolve — FFT-based wave kernel propagates information across the field
  3. Gather — tokens read back from the field at their positions

Each attention head learns three parameters: frequency, damping, and phase — controlling how information flows between tokens.

Why It Matters

Standard attention computes all token pairs — O(N²). At long context:

32K tokens:   Standard needs 35 GB    →  Wave Field needs 6.7 GB
128K tokens:  Standard OOMs           →  Wave Field runs at 26.7 GB
1M+ tokens:   Standard impossible     →  Wave Field feasible

Wave Field's throughput increases with context length while standard decreases.

Status

Active research. Training and benchmarking at 130M–1.5B parameter scale.

Author

Badaramoni Avinash

License

All rights reserved.

About

Wave Field AI — a efficient attention architecture for language models

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors