Hugging Face's logo

Spaces:
MatterWaveAI
/
README
Running

App Files Community
README
File size: 20,895 Bytes
822e652
5d10602
a2fe212
822e652
 
 
5d10602
822e652
 
5d10602
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
275cc3f
5d10602
7eff3ad
 
 
5d10602
 
 
 
 
275cc3f
5d10602
 
 
 
 
 
 
 
 
 
7eff3ad
 
 
5d10602
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
275cc3f
 
 
 
 
 
 
 
5d10602
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
---
title: README
emoji: 🌀
colorFrom: indigo
colorTo: blue
sdk: static
pinned: false
---

<style>
  .icon {
    width: 1.4em;
    height: 1.4em;
    vertical-align: -.35em;
    padding: .15em .45em .15em .15em;
  }
  ul.social {
    list-style: none;
    padding-left: 0;
  }
  ul.social li {
    padding-left: .25em;
    margin-bottom: .35em;
    display: flex;
    align-items: center;
  }
</style>

_**Purely Rotational Foundation Models & Geometric Optimization**_

**MatterWave AI** researches and builds **[Geometric Rotation Transformers (GRT)](https://huggingface.co/MatterWaveAI/grt-30m/resolve/main/GRT_Whitepaper.pdf)** — a foundation architecture that replaces dense linear weight matrices (`nn.Linear`) in attention (`Q, K, V, O`) and feed-forward (`expand, contract`) blocks with cascades of content-modulated 2×2 Givens rotations on `SO(D)` interleaved with butterfly permutations.

## Build and Scale with Geometric Rotation Transformers

Because every transform in the GRT residual path is an orthogonal rotation cascade, hidden-state norms are preserved by construction rather than by post-hoc rescaling. GRT checkpoints execute deterministically with bit-for-bit greedy generation parity across **PyTorch** (CUDA, ROCm, Intel XPU), **vLLM**, and **llama.cpp** (CPU AVX2, Vulkan, and ROCm HIP).

Starting from our pretrained **GRT-10M** and **GRT-20M** NeurIPS submission checkpoints (trained with first-generation **PairAdam**) and our 24-hour Qwen 3 distilled **GRT-30M** base checkpoint, replacing the optimizer with **HuygensSymplecticAdam** and training across a multi-round curriculum delivers substantial improvements across the 10-task **SLM-Bench** evaluation suite (13,144 held-out multiple-choice examples across BoolQ, ARC-Easy, ARC-Challenge, OpenBookQA, PIQA, WinoGrande, HellaSwag, QNLI, and CommonsenseQA, plus 500 held-out GSM8K mathematical reasoning problems):

* **[GRT-10M-Curriculum (8.7M total / 2.2M backbone)](https://huggingface.co/MatterWaveAI/grt-10m-curriculum)** — Lifts 9-task multiple-choice accuracy from **37.61%** (PairAdam NeurIPS Baseline) to **41.71%** (Single-Pass) to **45.74%** (Multi-Round Curriculum, **+8.13%**), while GSM8K 4-way verification reaches **96.60%** (perplexity reduced **5.77×** from 145.85 to **25.28**).
* **[GRT-20M-Curriculum (16.7M total / 10.3M backbone)](https://huggingface.co/MatterWaveAI/grt-20m-curriculum)** — Lifts 9-task multiple-choice accuracy from **38.63%** (PairAdam NeurIPS Baseline) to **45.03%** (Single-Pass) to **48.17%** (Multi-Round Curriculum, **+9.54%**), while GSM8K 4-way verification reaches **98.20%** (perplexity reduced **7.86×** from 60.12 to **7.65**).
* **[GRT-30M-Curriculum (29.3M total / 9.74M backbone)](https://huggingface.co/MatterWaveAI/grt-30m-curriculum)** — Lifts 9-task multiple-choice accuracy on the 24-hour Qwen 3 distilled base from **39.61%** (PairAdam Baseline) to **43.86%** (Single-Pass) to **48.07%** (Multi-Round Curriculum, **+8.46%**), while GSM8K 4-way verification reaches **99.20%** (perplexity reduced **3.68×** from 27.99 to **7.60**).

## System-1 Cognitive Telemetry & Multimodal Specialists

Beyond base causal language modeling, the **[GRT-30M](https://huggingface.co/MatterWaveAI/grt-30m)** backbone powers specialized reasoning routers, retrieval encoders, and unified multimodal checkpoints:

* **[Open-S1-GRT-30M](https://huggingface.co/MatterWaveAI/open-s1-grt-30m)** — 5-head System-1 cognitive telemetry router (action routing, escalation, loop guard, specification support, and verification gating) reaching **95.83%** 5-family macro accuracy (**92.00%** `MC_SHUFFLE_V2` transfer with `HuygensSymplecticAdam`; **ECE = 0.0095** with `SymplecticAdam`).
* **[GRT Architecture Whitepaper (PDF)](https://huggingface.co/MatterWaveAI/grt-30m/resolve/main/GRT_Whitepaper.pdf)** — Full mathematical formulation, kernel design, and multi-backend verification report.

## Explore the MatterWave AI Model Suite

Browse our open-weights repositories by category:

<div class="grid lg:grid-cols-2 gap-x-4 gap-y-7 p2">
  <div class="col-span-1 p2">
    <p><strong>Foundation & Curriculum LMs</strong></p>
    <ul class="social">
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M184 0c30.9 0 56 25.1 56 56l0 400c0 30.9-25.1 56-56 56c-28.9 0-52.7-21.9-55.7-50.1c-5.2 1.4-10.7 2.1-16.3 2.1c-35.3 0-64-28.7-64-64c0-7.4 1.3-14.6 3.6-21.2C21.4 367.4 0 338.2 0 304c0-31.9 18.7-59.5 45.8-72.3C37.1 220.8 32 207 32 192c0-30.7 21.6-56.3 50.4-62.6C80.8 123.9 80 118 80 112c0-29.9 20.6-55.1 48.3-62.1C131.3 21.9 155.1 0 184 0zM328 0c28.9 0 52.6 21.9 55.7 49.9c27.8 7 48.3 32.1 48.3 62.1c0 6-.8 11.9-2.4 17.4c28.8 6.2 50.4 31.9 50.4 62.6c0 15-5.1 28.8-13.8 39.7C493.3 244.5 512 272.1 512 304c0 34.2-21.4 63.4-51.6 74.8c2.3 6.6 3.6 13.8 3.6 21.2c0 35.3-28.7 64-64 64c-5.6 0-11.1-.7-16.3-2.1c-3 28.2-26.8 50.1-55.7 50.1c-30.9 0-56-25.1-56-56l0-400c0-30.9 25.1-56 56-56z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-curriculum">GRT-30M-Curriculum (48.07% MC / 99.2% GSM8K)</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M184 0c30.9 0 56 25.1 56 56l0 400c0 30.9-25.1 56-56 56c-28.9 0-52.7-21.9-55.7-50.1c-5.2 1.4-10.7 2.1-16.3 2.1c-35.3 0-64-28.7-64-64c0-7.4 1.3-14.6 3.6-21.2C21.4 367.4 0 338.2 0 304c0-31.9 18.7-59.5 45.8-72.3C37.1 220.8 32 207 32 192c0-30.7 21.6-56.3 50.4-62.6C80.8 123.9 80 118 80 112c0-29.9 20.6-55.1 48.3-62.1C131.3 21.9 155.1 0 184 0zM328 0c28.9 0 52.6 21.9 55.7 49.9c27.8 7 48.3 32.1 48.3 62.1c0 6-.8 11.9-2.4 17.4c28.8 6.2 50.4 31.9 50.4 62.6c0 15-5.1 28.8-13.8 39.7C493.3 244.5 512 272.1 512 304c0 34.2-21.4 63.4-51.6 74.8c2.3 6.6 3.6 13.8 3.6 21.2c0 35.3-28.7 64-64 64c-5.6 0-11.1-.7-16.3-2.1c-3 28.2-26.8 50.1-55.7 50.1c-30.9 0-56-25.1-56-56l0-400c0-30.9 25.1-56 56-56z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-20m-curriculum">GRT-20M-Curriculum (48.17% MC / 98.2% GSM8K)</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M184 0c30.9 0 56 25.1 56 56l0 400c0 30.9-25.1 56-56 56c-28.9 0-52.7-21.9-55.7-50.1c-5.2 1.4-10.7 2.1-16.3 2.1c-35.3 0-64-28.7-64-64c0-7.4 1.3-14.6 3.6-21.2C21.4 367.4 0 338.2 0 304c0-31.9 18.7-59.5 45.8-72.3C37.1 220.8 32 207 32 192c0-30.7 21.6-56.3 50.4-62.6C80.8 123.9 80 118 80 112c0-29.9 20.6-55.1 48.3-62.1C131.3 21.9 155.1 0 184 0zM328 0c28.9 0 52.6 21.9 55.7 49.9c27.8 7 48.3 32.1 48.3 62.1c0 6-.8 11.9-2.4 17.4c28.8 6.2 50.4 31.9 50.4 62.6c0 15-5.1 28.8-13.8 39.7C493.3 244.5 512 272.1 512 304c0 34.2-21.4 63.4-51.6 74.8c2.3 6.6 3.6 13.8 3.6 21.2c0 35.3-28.7 64-64 64c-5.6 0-11.1-.7-16.3-2.1c-3 28.2-26.8 50.1-55.7 50.1c-30.9 0-56-25.1-56-56l0-400c0-30.9 25.1-56 56-56z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-10m-curriculum">GRT-10M-Curriculum (45.74% MC / 96.6% GSM8K)</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 448 512" class="icon"><path fill="currentColor" d="M448 80l0 48c0 44.2-100.3 80-224 80S0 172.2 0 128L0 80C0 35.8 100.3 0 224 0S448 35.8 448 80zM393.2 214.7c20.8-7.4 39.9-16.9 54.8-28.6L448 288c0 44.2-100.3 80-224 80S0 332.2 0 288L0 186.1c14.9 11.8 34 21.2 54.8 28.6C99.7 230.7 159.5 240 224 240s124.3-9.3 169.2-25.3zM0 346.1c14.9 11.8 34 21.2 54.8 28.6C99.7 390.7 159.5 400 224 400s124.3-9.3 169.2-25.3c20.8-7.4 39.9-16.9 54.8-28.6l0 85.9c0 44.2-100.3 80-224 80S0 476.2 0 432l0-85.9z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m">GRT-30M (Qwen 3 Distilled Base + GGUF)</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 448 512" class="icon"><path fill="currentColor" d="M448 80l0 48c0 44.2-100.3 80-224 80S0 172.2 0 128L0 80C0 35.8 100.3 0 224 0S448 35.8 448 80zM393.2 214.7c20.8-7.4 39.9-16.9 54.8-28.6L448 288c0 44.2-100.3 80-224 80S0 332.2 0 288L0 186.1c14.9 11.8 34 21.2 54.8 28.6C99.7 230.7 159.5 240 224 240s124.3-9.3 169.2-25.3zM0 346.1c14.9 11.8 34 21.2 54.8 28.6C99.7 390.7 159.5 400 224 400s124.3-9.3 169.2-25.3c20.8-7.4 39.9-16.9 54.8-28.6l0 85.9c0 44.2-100.3 80-224 80S0 476.2 0 432l0-85.9z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-20m">GRT-20M</a> &nbsp;|&nbsp; <a href="https://huggingface.co/MatterWaveAI/grt-10m">GRT-10M</a></li>
    </ul>
  </div>
  <div class="col-span-1 p2">
    <p><strong>Reasoning, Routing & NLP Specialists</strong></p>
    <ul class="social">
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M176 24c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40c-35.3 0-64 28.7-64 64l-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0 0 56-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0 0 56-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0c0 35.3 28.7 64 64 64l0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40 56 0 0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40 56 0 0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40c35.3 0 64-28.7 64-64l40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0 0-56 40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0 0-56 40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0c0-35.3-28.7-64-64-64l0-40c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40-56 0 0-40c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40-56 0 0-40zM160 128l192 0c17.7 0 32 14.3 32 32l0 192c0 17.7-14.3 32-32 32l-192 0c-17.7 0-32-14.3-32-32l0-192c0-17.7 14.3-32 32-32zm192 32l-192 0 0 192 192 0 0-192z"/></svg><a href="https://huggingface.co/MatterWaveAI/open-s1-grt-30m">Open-S1-GRT-30M (System-1 Router)</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M176 24c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40c-35.3 0-64 28.7-64 64l-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0 0 56-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0 0 56-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0c0 35.3 28.7 64 64 64l0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40 56 0 0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40 56 0 0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40c35.3 0 64-28.7 64-64l40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0 0-56 40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0 0-56 40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0c0-35.3-28.7-64-64-64l0-40c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40-56 0 0-40c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40-56 0 0-40zM160 128l192 0c17.7 0 32 14.3 32 32l0 192c0 17.7-14.3 32-32 32l-192 0c-17.7 0-32-14.3-32-32l0-192c0-17.7 14.3-32 32-32zm192 32l-192 0 0 192 192 0 0-192z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-chat">GRT-30M-Chat</a> &nbsp;|&nbsp; <a href="https://huggingface.co/MatterWaveAI/grt-30m-mor">GRT-30M-MoR</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M176 24c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40c-35.3 0-64 28.7-64 64l-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0 0 56-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0 0 56-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0c0 35.3 28.7 64 64 64l0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40 56 0 0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40 56 0 0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40c35.3 0 64-28.7 64-64l40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0 0-56 40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0 0-56 40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0c0-35.3-28.7-64-64-64l0-40c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40-56 0 0-40c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40-56 0 0-40zM160 128l192 0c17.7 0 32 14.3 32 32l0 192c0 17.7-14.3 32-32 32l-192 0c-17.7 0-32-14.3-32-32l0-192c0-17.7 14.3-32 32-32zm192 32l-192 0 0 192 192 0 0-192z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-code">GRT-30M-Code</a> &nbsp;|&nbsp; <a href="https://huggingface.co/MatterWaveAI/grt-30m-prm">GRT-30M-PRM</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M176 24c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40c-35.3 0-64 28.7-64 64l-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0 0 56-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0 0 56-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0c0 35.3 28.7 64 64 64l0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40 56 0 0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40 56 0 0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40c35.3 0 64-28.7 64-64l40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0 0-56 40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0 0-56 40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0c0-35.3-28.7-64-64-64l0-40c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40-56 0 0-40c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40-56 0 0-40zM160 128l192 0c17.7 0 32 14.3 32 32l0 192c0 17.7-14.3 32-32 32l-192 0c-17.7 0-32-14.3-32-32l0-192c0-17.7 14.3-32 32-32zm192 32l-192 0 0 192 192 0 0-192z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-qa">GRT-30M-QA</a> &nbsp;|&nbsp; <a href="https://huggingface.co/MatterWaveAI/grt-30m-ner">GRT-30M-NER</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M176 24c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40c-35.3 0-64 28.7-64 64l-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0 0 56-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0 0 56-40 0c-13.3 0-24 10.7-24 24s10.7 24 24 24l40 0c0 35.3 28.7 64 64 64l0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40 56 0 0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40 56 0 0 40c0 13.3 10.7 24 24 24s24-10.7 24-24l0-40c35.3 0 64-28.7 64-64l40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0 0-56 40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0 0-56 40 0c13.3 0 24-10.7 24-24s-10.7-24-24-24l-40 0c0-35.3-28.7-64-64-64l0-40c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40-56 0 0-40c0-13.3-10.7-24-24-24s-24 10.7-24 24l0 40-56 0 0-40zM160 128l192 0c17.7 0 32 14.3 32 32l0 192c0 17.7-14.3 32-32 32l-192 0c-17.7 0-32-14.3-32-32l0-192c0-17.7 14.3-32 32-32zm192 32l-192 0 0 192 192 0 0-192z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-lora">GRT-30M-LoRA</a></li>
    </ul>
  </div>
  <div class="col-span-1 p2">
    <p><strong>Retrieval, Embeddings & Tokenizer</strong></p>
    <ul class="social">
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 640 512" class="icon"><path fill="currentColor" d="M175 389.4c-9.8 16-15 34.3-15 53.1c-10 3.5-20.8 5.5-32 5.5c-53 0-96-43-96-96L32 64C14.3 64 0 49.7 0 32S14.3 0 32 0L96 0l64 0 64 0c17.7 0 32 14.3 32 32s-14.3 32-32 32l0 245.9-49 79.6zM96 64l0 96 64 0 0-96L96 64zM352 0L480 0l32 0c17.7 0 32 14.3 32 32s-14.3 32-32 32l0 150.9L629.7 406.2c6.7 10.9 10.3 23.5 10.3 36.4c0 38.3-31.1 69.4-69.4 69.4l-309.2 0c-38.3 0-69.4-31.1-69.4-69.4c0-12.8 3.6-25.4 10.3-36.4L320 214.9 320 64c-17.7 0-32-14.3-32-32s14.3-32 32-32l32 0zm32 64l0 160c0 5.9-1.6 11.7-4.7 16.8L330.5 320l171 0-48.8-79.2c-3.1-5-4.7-10.8-4.7-16.8l0-160-64 0z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-embedding">GRT-30M-Embedding (Circular-Pooled)</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 640 512" class="icon"><path fill="currentColor" d="M175 389.4c-9.8 16-15 34.3-15 53.1c-10 3.5-20.8 5.5-32 5.5c-53 0-96-43-96-96L32 64C14.3 64 0 49.7 0 32S14.3 0 32 0L96 0l64 0 64 0c17.7 0 32 14.3 32 32s-14.3 32-32 32l0 245.9-49 79.6zM96 64l0 96 64 0 0-96L96 64zM352 0L480 0l32 0c17.7 0 32 14.3 32 32s-14.3 32-32 32l0 150.9L629.7 406.2c6.7 10.9 10.3 23.5 10.3 36.4c0 38.3-31.1 69.4-69.4 69.4l-309.2 0c-38.3 0-69.4-31.1-69.4-69.4c0-12.8 3.6-25.4 10.3-36.4L320 214.9 320 64c-17.7 0-32-14.3-32-32s14.3-32 32-32l32 0zm32 64l0 160c0 5.9-1.6 11.7-4.7 16.8L330.5 320l171 0-48.8-79.2c-3.1-5-4.7-10.8-4.7-16.8l0-160-64 0z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-reranker">GRT-30M-Reranker</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 640 512" class="icon"><path fill="currentColor" d="M175 389.4c-9.8 16-15 34.3-15 53.1c-10 3.5-20.8 5.5-32 5.5c-53 0-96-43-96-96L32 64C14.3 64 0 49.7 0 32S14.3 0 32 0L96 0l64 0 64 0c17.7 0 32 14.3 32 32s-14.3 32-32 32l0 245.9-49 79.6zM96 64l0 96 64 0 0-96L96 64zM352 0L480 0l32 0c17.7 0 32 14.3 32 32s-14.3 32-32 32l0 150.9L629.7 406.2c6.7 10.9 10.3 23.5 10.3 36.4c0 38.3-31.1 69.4-69.4 69.4l-309.2 0c-38.3 0-69.4-31.1-69.4-69.4c0-12.8 3.6-25.4 10.3-36.4L320 214.9 320 64c-17.7 0-32-14.3-32-32s14.3-32 32-32l32 0zm32 64l0 160c0 5.9-1.6 11.7-4.7 16.8L330.5 320l171 0-48.8-79.2c-3.1-5-4.7-10.8-4.7-16.8l0-160-64 0z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-tokenizer">GRT-30M-Tokenizer & Polar Bank</a></li>
    </ul>
  </div>
  <div class="col-span-1 p2">
    <p><strong>Multimodal (Omni, Vision, Speech, Audio)</strong></p>
    <ul class="social">
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M208.5 62.3l28.1-36.9C248.8 9.4 267.8 0 288 0c28.5 0 53.6 18.7 61.8 46l17.8 59.4c10.3 34.4 36.1 62 69.8 74.6l39.8 14.9c20.9 7.9 34.8 27.9 34.8 50.2c0 16.9-7.9 32.8-21.5 42.9l-67.3 50.5c-24.3 18.2-37.2 47.9-33.8 78.1l2.5 22.7c4.3 38.7-26 72.6-65 72.6c-14.8 0-29.3-5.1-40.8-14.3l-55.4-44.3c-4.5-3.6-9.3-6.7-14.5-9.2c-15.8-7.9-33.7-10.4-51-7.3L82.4 451.9C47.8 458.2 16 431.6 16 396.5c0-13.2 4.7-26 13.1-36.2l11.2-13.4c14.6-17.4 22.6-39.4 22.6-62.1c0-18.8-5.5-37.2-15.8-53L8.8 173.5C3.1 164.7 0 154.4 0 143.9c0-33.4 30.1-58.8 63-53.2l51.3 8.7c35.9 6.1 72.2-8.2 94.2-37.1z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-omni">GRT-30M-Omni (Vision + Audio + Speech + Text)</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M208.5 62.3l28.1-36.9C248.8 9.4 267.8 0 288 0c28.5 0 53.6 18.7 61.8 46l17.8 59.4c10.3 34.4 36.1 62 69.8 74.6l39.8 14.9c20.9 7.9 34.8 27.9 34.8 50.2c0 16.9-7.9 32.8-21.5 42.9l-67.3 50.5c-24.3 18.2-37.2 47.9-33.8 78.1l2.5 22.7c4.3 38.7-26 72.6-65 72.6c-14.8 0-29.3-5.1-40.8-14.3l-55.4-44.3c-4.5-3.6-9.3-6.7-14.5-9.2c-15.8-7.9-33.7-10.4-51-7.3L82.4 451.9C47.8 458.2 16 431.6 16 396.5c0-13.2 4.7-26 13.1-36.2l11.2-13.4c14.6-17.4 22.6-39.4 22.6-62.1c0-18.8-5.5-37.2-15.8-53L8.8 173.5C3.1 164.7 0 154.4 0 143.9c0-33.4 30.1-58.8 63-53.2l51.3 8.7c35.9 6.1 72.2-8.2 94.2-37.1z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-vl">GRT-30M-VL</a> &nbsp;|&nbsp; <a href="https://huggingface.co/MatterWaveAI/grt-30m-image">GRT-30M-Image</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M208.5 62.3l28.1-36.9C248.8 9.4 267.8 0 288 0c28.5 0 53.6 18.7 61.8 46l17.8 59.4c10.3 34.4 36.1 62 69.8 74.6l39.8 14.9c20.9 7.9 34.8 27.9 34.8 50.2c0 16.9-7.9 32.8-21.5 42.9l-67.3 50.5c-24.3 18.2-37.2 47.9-33.8 78.1l2.5 22.7c4.3 38.7-26 72.6-65 72.6c-14.8 0-29.3-5.1-40.8-14.3l-55.4-44.3c-4.5-3.6-9.3-6.7-14.5-9.2c-15.8-7.9-33.7-10.4-51-7.3L82.4 451.9C47.8 458.2 16 431.6 16 396.5c0-13.2 4.7-26 13.1-36.2l11.2-13.4c14.6-17.4 22.6-39.4 22.6-62.1c0-18.8-5.5-37.2-15.8-53L8.8 173.5C3.1 164.7 0 154.4 0 143.9c0-33.4 30.1-58.8 63-53.2l51.3 8.7c35.9 6.1 72.2-8.2 94.2-37.1z"/></svg><a href="https://huggingface.co/MatterWaveAI/grt-30m-asr">GRT-30M-ASR</a> &nbsp;|&nbsp; <a href="https://huggingface.co/MatterWaveAI/grt-30m-tts">GRT-30M-TTS</a></li>
    </ul>
  </div>
  <div class="col-span-1 p2">
    <p><strong>Rotational JEPA, SSM &amp; RNN World Models (Preview)</strong></p>
    <ul class="social">
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M256 512A256 256 0 1 0 256 0a256 256 0 1 0 0 512zM369 209L241 337c-9.4 9.4-24.6 9.4-33.9 0l-64-64c-9.4-9.4-9.4-24.6 0-33.9s24.6-9.4 33.9 0l47 47L335 175c9.4-9.4 24.6-9.4 33.9 0s9.4 24.6 0 33.9z"/></svg><a href="https://huggingface.co/MatterWaveAI/helmholtz-1">HELMHOLTZ-1 (Rotational JEPA World Model)</a> &nbsp;|&nbsp; <a href="https://huggingface.co/MatterWaveAI/helmholtz-1-omni">HELMHOLTZ-1-Omni</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M256 512A256 256 0 1 0 256 0a256 256 0 1 0 0 512zM369 209L241 337c-9.4 9.4-24.6 9.4-33.9 0l-64-64c-9.4-9.4-9.4-24.6 0-33.9s24.6-9.4 33.9 0l47 47L335 175c9.4-9.4 24.6-9.4 33.9 0s9.4 24.6 0 33.9z"/></svg><a href="https://huggingface.co/MatterWaveAI/helmholtz-mamba-1">HELMHOLTZ-Mamba-1 (Rotational SSM)</a> &nbsp;|&nbsp; <a href="https://huggingface.co/MatterWaveAI/helmholtz-rnn-1">HELMHOLTZ-RNN-1 (Unitary RNN)</a></li>
      <li><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" class="icon"><path fill="currentColor" d="M256 512A256 256 0 1 0 256 0a256 256 0 1 0 0 512zM369 209L241 337c-9.4 9.4-24.6 9.4-33.9 0l-64-64c-9.4-9.4-9.4-24.6 0-33.9s24.6-9.4 33.9 0l47 47L335 175c9.4-9.4 24.6-9.4 33.9 0s9.4 24.6 0 33.9z"/></svg><a href="https://huggingface.co/MatterWaveAI/s1-grt-jepa">S1-GRT-JEPA</a> &nbsp;|&nbsp; <a href="https://huggingface.co/MatterWaveAI/s1-grt-mamba">S1-GRT-Mamba</a> &nbsp;|&nbsp; <a href="https://huggingface.co/MatterWaveAI/s1-grt-rnn">S1-GRT-RNN</a></li>
    </ul>
  </div>
</div>