svs/README.md

# Subspace Validity Suite (SVS)

**A diagnostic toolkit for validating claimed "visual directions" in Vision-Language Models.**

> Do your visual subspace directions actually capture visual content,
> or are they just network geometry?

SVS provides 6 tests that any claimed visual direction must pass before
being attributed mechanistic significance.

## Quick Start

```python
from svs_toolkit import SubspaceValiditySuite

svs = SubspaceValiditySuite()

# Run all tests
report = svs.full_report(
    directions=your_visual_directions,      # (k, d) numpy array
    h_visual=visual_hidden_states,          # list of (d,) arrays
    h_gibberish=gibberish_hidden_states,    # list of (d,) arrays
    h_factual=factual_hidden_states,        # optional
    h_math=math_hidden_states,              # optional
)

svs.print_report(report)
```

## The 6 Tests

| # | Test | What it checks | PASS condition |
|---|------|---------------|----------------|
| 1 | Gibberish Specificity | Do directions respond more to visual than gibberish? | Gib/Vis < 0.8, TOST rejects equivalence |
| 2 | Cross-Type Discrimination | Can projection magnitude tell visual from non-visual? | AUROC > 0.65 |
| 3 | Projection Magnitude | Do directions capture more variance than random? | Ratio > 1.5x |
| 4 | Anisotropy Orthogonality | Are directions different from general PCA? | Mean cosine < 0.3 |
| 5 | Direction Consistency | Are directions stable across calibration splits? | Split alignment > 0.8 |
| 6 | Pairwise Discrimination | Do projections encode visual content specifically? | Only visual pairs separable |

## Key Finding

We tested 11 methods across 3 VLM architectures, 2 text-only backbones,
and 2 non-VLM transformer families. **Every method fails the SVS.**

PCA-based "visual directions" capture generic network geometry — the
dominant modes of the weight matrices — not visual content. Gibberish
activates them identically to visual descriptions (Gib/Vis ≈ 1.00,
TOST equivalence p < 0.0001, N = 4,000).

## Installation

```bash
pip install numpy scipy scikit-learn
```

No additional dependencies required. Single file: `svs_toolkit.py`.

## Generating Test Inputs

SVS needs hidden states from a text-only backbone (e.g., Vicuna-7B):

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model = AutoModelForCausalLM.from_pretrained("lmsys/vicuna-7b-v1.5",
    torch_dtype=torch.float16, device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("lmsys/vicuna-7b-v1.5")

def get_hidden(prompt, layer=16):
    inp = tokenizer(prompt, return_tensors="pt").to(model.device)
    with torch.no_grad():
        out = model(**inp, output_hidden_states=True)
    return out.hidden_states[layer][0, -1, :].cpu().float().numpy()

# Visual prompts
h_visual = [get_hidden(p) for p in [
    "A kitchen with a table and chairs.",
    "A beach with surfers and umbrellas.",
    # ... 20+ prompts
]]

# Gibberish prompts
h_gibberish = [get_hidden(p) for p in [
    "Xkq plm wvt zzz brrn.",
    "Qwzyx nkl jjj hhh ttttt.",
    # ... 20+ prompts
]]
```

## Citation

```bibtex
@inproceedings{svs2027,
  title={The Subspace Validity Suite: Do Visual Directions in
         Vision-Language Models Survive Basic Sanity Checks?},
  author={Anonymous},
  booktitle={WACV},
  year={2027}
}
```

## License

MIT