Tighten intro: compact first sentence, drop em-dash clause, reword

README.md

@@ -29,14 +29,13 @@ actuation, and memory-aware decision-making in a single architecture. RLDX-1
 achieves state-of-the-art performance across diverse simulation benchmarks
 and is fully validated on real-world hardware.
 
-This repository hosts **`RLDX-1-PT-IMG`** — a lightweight, **image-input**
-variant of the `RLDX-1-PT` foundation checkpoint. Whereas `RLDX-1-PT` defaults
-to **video input (4 frames per camera view)**, `RLDX-1-PT-IMG` consumes a
-**single RGB image per view** (`video_length=1`). This trades a **minimal drop in success rate** for a **substantially lighter and faster** policy —
-fewer vision tokens, lower latency, and reduced memory — making it well suited
-to real-time and resource-constrained deployment. It is pretrained on the same
-broad mixture of public manipulation corpora and serves as a starting point
-for new embodiments and tasks.
+This repository hosts **`RLDX-1-PT-IMG`**: a lightweight, **image-input**
+variant of the `RLDX-1-PT`, which uses 4-frame video inputs. This trades a
+**minimal drop in success rate** for a **substantially lighter and faster**
+policy, making it well suited to real-time and resource-constrained
+deployment. It is pre-trained on the same broad mixture of public manipulation
+corpora, providing a lightweight starting point for rapid experimentation on
+new embodiments and tasks.
 
 <p align="center">
 <img src="architecture.png" width="90%" alt="RLDX-1 architecture">