docs(model-card): replace raw LaTeX with clean Unicode and code formatting

README.md

@@ -22,7 +22,7 @@ pipeline_tag: text-generation
 
 # GRT-30M (24-Hour Qwen 3 Distillation Preview)
 
-A **Geometric Rotation Transformer (`GRTForCausalLM`)** distilled from **Qwen 3** (sharing the `151,669`-token Qwen 3 BPE vocabulary) in a single **24-hour distillation run** (`10.79B` tokens at `~123.9k tok/s` across `8x NVIDIA A100-SXM4-80GB`). GRT replaces every dense linear projection (`nn.Linear`) in attention (`Q, K, V, O`) and feed-forward (`expand, contract`) with cascades of content-modulated $2 \times 2$ Givens rotations on $\mathrm{SO}(D)$ interleaved with butterfly permutations.
+A **Geometric Rotation Transformer (`GRTForCausalLM`)** distilled from **Qwen 3** (sharing the `151,669`-token Qwen 3 BPE vocabulary) in a single **24-hour distillation run** (`10.79B` tokens at `~123.9k tok/s` across `8x NVIDIA A100-SXM4-80GB`). GRT replaces every dense linear projection (`nn.Linear`) in attention (`Q, K, V, O`) and feed-forward (`expand, contract`) with cascades of content-modulated `2×2` Givens rotations on `SO(D)` interleaved with butterfly permutations.
 
 **This is a base research checkpoint distilled from Qwen 3, not an instruction-tuned assistant.** It demonstrates that a rotation-only student backbone with just **`9.74M` non-embedding parameters** (`29.31M` total stored parameters with the 151,669-token Qwen 3 vocabulary):
 1. Distills stably from Qwen 3 at **`123,900 tok/s`** to **`val_loss = 4.4776`** (`val_ppl = 88.02`, **`27.2%` top-1 token accuracy** over a `151,669`-token vocabulary).
@@ -123,7 +123,7 @@ Evaluated zero-shot (no task fine-tuning) via teacher-forced length-normalized l
 
 #### PyTorch ROCm (`gfx1151`, `BF16`, Fused HIP Kernels)
 
-| Workload | Batch $\times$ Seq (`B x T`) | Latency (`ms`) | Throughput (`tok/s`) |
+| Workload | Batch × Seq (`B × T`) | Latency (`ms`) | Throughput (`tok/s`) |
 |---|---|---:|---:|
 | **Prefill (`B=1, T=512`)** | `1 x 512` | `129.36` | `3,957.9` |
 | **Prefill (`B=1, T=2048`)** | `1 x 2048` | `357.89` | **`5,722.5`** |
@@ -177,16 +177,19 @@ sha256sum --ignore-missing -c SHA256SUMS
 
 ## Architecture Overview
 
-GRT replaces dense weight matrices (`nn.Linear`) in both attention and FFN blocks with cascades of learned $2 \times 2$ Givens rotations arranged in a butterfly topology. At each stage $k$, pairs of coordinates $(x_0, x_1)$ are rotated by an angle $\theta_k$ modulated by 4-component $\mathrm{Cl}(2,0)$ multivector features:
+GRT replaces dense weight matrices (`nn.Linear`) in both attention and FFN blocks with cascades of learned `2×2` Givens rotations arranged in a butterfly topology. At each stage `k`, pairs of coordinates `(x0, x1)` are rotated by an angle `theta_k` modulated by 4-component `Cl(2,0)` multivector features:
 
-$$\mathrm{feat}(x_0, x_1) = \bigl[x_0^2 + x_1^2,\; x_0,\; x_1,\; x_0 x_{1,\text{next}} - x_1 x_{0,\text{next}}\bigr]$$
-
-$$\theta_k = \alpha_k + \tanh\!\bigl(\widehat{\mathrm{feat}}(x_0, x_1) \cdot C_k\bigr)$$
+```text
+feat(x0, x1) = [x0^2 + x1^2,  x0,  x1,  x0 * x1_next - x1 * x0_next]
+theta_k      = alpha_k + tanh(norm(feat(x0, x1)) @ C_k)
+r0           = x0 * cos(theta_k) - x1 * sin(theta_k)
+r1           = x0 * sin(theta_k) + x1 * cos(theta_k)
+```
 
 - **`GRTEmbedding` & `GRTCosineHead`** — Token embeddings are stored in polar form (`phase_angles` + `log_magnitude`) and tied to a scaled cosine-similarity language modeling head (`lm_head.output_scale = 40.23`).
 - **`GRTAttention`** — 11-stage content-modulated Givens butterfly cascades for `Q, K, V, O`, paired with standard scaled dot-product attention and learnable RoPE frequencies.
 - **`GRTMixer` & `GRTFFN`** — Cross-head Givens mixing stage followed by 21-stage rotational FFN expansion/contraction and phase-preserving `GRTRadialGELU`.
-- **`GRTPearlNorm`** — Pair-aware RMS normalization that normalizes pair radii $\sqrt{x_{2i}^2 + x_{2i+1}^2}$ while preserving phase angles.
+- **`GRTPearlNorm`** — Pair-aware RMS normalization that normalizes pair radii `sqrt(x_{2i}^2 + x_{2i+1}^2)` while preserving phase angles.
 
 ---
 

SHA256SUMS

@@ -2,6 +2,6 @@
 5dbb8b936d1e2411f30d1ecf072a4d7c2f2df6e299b0d11c7ef5cbaf72ebd5bc  generation_config.json
 65bda04ea57b1799e7b0c6d09102475a81c3247c2e29fdb5cbdcdf6d52a21af2  grt-30m-f32.gguf
 205f76f3ed59fee0ed54bb8133ea1b1a64527994c68776066057b4707f1b2456  model.safetensors
-c4855c6e2a1dd0d2877525588da601b924e9dde79836f7491423f88309cbd819  README.md
+025b49eeb0c4dab8e792d4931140b7334191bbde8f0d9795b091017cbdfff14c  README.md
 be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506  tokenizer.json
 5f007d04324afa93876e2a484190e868136c69dfe0e1a29a1063eae583df9d32  tokenizer_config.json