Spaces:
Sleeping
Sleeping
File size: 1,279 Bytes
0773bfe 26ee859 0773bfe 26ee859 d1f8780 0773bfe 26ee859 0773bfe 26ee859 | 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 | ---
title: SRT-NLA v1 demo
emoji: 🪞
colorFrom: indigo
colorTo: green
sdk: gradio
sdk_version: 4.44.1
python_version: "3.10.13"
app_file: app.py
pinned: false
license: apache-2.0
hardware: zero-a10g
short_description: Latent autoencoder over Qwen2.5-7B L20
models:
- RiverRider/srt-nla-av-v1
- Qwen/Qwen2.5-7B
---
# SRT-NLA v1 — latent autoencoder demo
Two interactive views of the `RiverRider/srt-nla-av-v1` verbalizer.
1. **Round-trip autoencoder.** Encode a passage with frozen Qwen2.5-7B at L20,
decode the resulting 3584-dim vector with the 12.7M-parameter NLA verbalizer,
and re-encode the verbalization to measure round-trip fidelity
(`centered fve_nrm`, `rho_norm`). Best-of-N rerank is included — this is the
technique that takes greedy `ρ ≈ 0.26` up to `ρ ≈ 0.92` in the paper.
2. **Latent arithmetic.** Encode two passages, slide an interpolation `α`, and
verbalize the mixture vector.
## Anchors (paper §3)
| Anchor | centered fve_nrm |
|---|---|
| Random floor | 0.510 |
| NN-retrieval over pool=2000 | 0.71 |
| Qwen paraphrase ceiling | 0.799 |
`rho_norm = (cen − 0.510) / 0.289` linearly remaps to `[0, 1]` against these
anchors.
## Source
Code: https://github.com/space-bacon/SRT (branch `nla`, `docs/hf/nla_v1_demo/`).
|