Update README.md

README.md

@@ -1,4 +1,4 @@
----
+[[---
 language: en
 license: mit
 library_name: open-synth-miner
@@ -18,15 +18,16 @@ datasets:
 pipeline_tag: time-series-forecasting
 ---
 
-# PAG — Hybrid SN50 Probabilistic Price Forecaster
+# SN50-Hybrid-Hub — Model Repository
+
 
-**PAG** is a hybrid probabilistic price forecasting model built for [Bittensor Subnet 50 (Synth)](https://bittensor.com/). It generates Monte Carlo price path simulations scored by **CRPS** (Continuous Ranked Probability Score) — a metric that rewards well-calibrated probability distributions, not just point accuracy.
+**SN50-Hybrid-Hub — Model Repository** is a hybrid probabilistic price forecasting model suite built for [Bittensor Subnet 50 (Synth)](https://bittensor.com/). It generates Monte Carlo price path simulations scored by **CRPS** (Continuous Ranked Probability Score) — a metric that rewards well-calibrated probability distributions, not just point accuracy.
 
 Built with [open-synth-miner](https://github.com/tensorlink-dev/open-synth-miner) and trained/published via [synth-city](https://github.com/tensorlink-dev/synth-city).
 
 ## Model Description
 
-PAG uses a **hybrid backbone architecture** — stacking multiple neural network paradigms (e.g. attention + recurrence, convolution + frequency analysis) into a single composable pipeline. All backbone blocks share a uniform tensor interface:
+SN50-Hybrid-Hub — Model Repository uses a **hybrid backbone architecture** — stacking multiple neural network paradigms (e.g. attention + recurrence, convolution + frequency analysis) into a single composable pipeline. All backbone blocks share a uniform tensor interface:
 
 ```
 (batch, seq, d_model) → (batch, seq, d_model)
@@ -259,4 +260,4 @@ If you use this model or the underlying framework, please reference:
 
 ## License
 
-MIT
+MIT](https://x.com/tensorlink_dev/status/2028323835393753118)](https://x.com/tensorlink_dev/status/2028323835393753118)