Update README.md

README.md

@@ -1,5 +1,10 @@
+---
+license: mit
+datasets:
+- EleutherAI/fineweb-edu-dedup-10b
+base_model:
+- deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
+---
 These transcoders were trained on the outputs of the first 15 MLPs in deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B. We used 10 billion tokens from FineWeb edu deduped at a context length of 2048. The number of latents is 65,536 and a linear skip connection is included.
 
-Fraction of variance unexplained ranges from 0.01 to 0.37.
-
-The fineweb-edu-dedup sample is EleutherAI/fineweb-edu-dedup-10b.
+Fraction of variance unexplained ranges from 0.01 to 0.37.