Finalize model card: SLM terminology, EdgeJSON v3.1 clarification, version comparison context

README.md

@@ -17,13 +17,13 @@ pipeline_tag: text-generation
 
 # Maaza SLM-360M-JSON v1.2
 
-**First sub-400M model with consistent complex schema wins on EdgeJSON benchmark.**
+**First SLM with complex schema wins on EdgeJSON v3.1 normalized benchmark.**
 
 A 360M parameter model fine-tuned for high-accuracy JSON extraction. v1.2 introduces extended context (2048 tokens) for improved complex schema handling.
 
 ## Performance
 
-### EdgeJSON v3 Benchmark (Normalized Scoring)
+### EdgeJSON v3.1 Benchmark (Normalized Scoring)
 
 | Metric | Score |
 |--------|-------|
@@ -41,7 +41,7 @@ A 360M parameter model fine-tuned for high-accuracy JSON extraction. v1.2 introd
 
 ### Version Comparison
 
-> **Note:** v1.1+ use EdgeJSON v3.1 normalized scoring (case-insensitive keys). This is stricter on simple/medium schemas but fairer overall. v1.2 is the first sub-400M model to solve complex schemas under the standardized benchmark.
+> **Note:** v1.1+ use EdgeJSON v3.1 normalized scoring (case-insensitive keys). This is stricter on simple/medium schemas but fairer overall. v1.2 is the first SLM to solve complex schemas under the standardized benchmark.
 
 | Version | EdgeJSON Eval | Overall (JSONExact) | Complex | Notes |
 |---------|---------------|---------------------|---------|-------|
@@ -147,7 +147,7 @@ python benchmarks/edge_json/scripts/eval.py \
 
 ```bibtex
 @misc{cyclecore2025maaza,
-  title={Maaza SLM-360M-JSON: Sub-400M JSON Extraction},
+  title={Maaza SLM-360M-JSON: Small Language Model for JSON Extraction},
   author={CycleCore Technologies},
   year={2025},
   howpublished={\url{https://huggingface.co/CycleCoreTechnologies/Maaza-SLM-360M-JSON-v1.2}}