Update README.md

README.md

@@ -15,13 +15,15 @@ base_model: roberta-base
 
 # DiSCo Situation-Entity Classifier (18-way)
 
-A `roberta-base` model fine-tuned to label an English **clause** with one of **18 Situation Entity (SE) types**, an extension of Smith's (2003) discourse-mode framework with Grisot (2018) boundedness. Each SE label decomposes into three content-agnostic attributes used in downstream analyses:
+A `roberta-base` model pre-trained on the SitEnt corpus (Friedrich, 2016) and then fine-tuned on the DiSCo corpus (Hemmatian, 2022; fortchoming) to label an English Elementary Discourse Unit (roughly a **clause**) with one of **18 Situation Entity (SE) types**. The labels are an extension of Smith's (2003) discourse-mode framework with Grisot (2018) boundedness (see Hemmatian, 2022, for details). Each SE label decomposes into three content-agnostic attributes used in downstream analyses:
 
 - **Genericity** of the main referent: `specific` / `generic`
 - **Eventivity** of the main verb constellation: `dynamic` / `stative`
 - **Boundedness/Habituality** of the eventuality: `static` / `episodic` / `habitual`
 
-This is the SE-classification component of the DiSCo pipeline (the companion model is [`BabakScrapes/disco-clause-segmenter`](https://huggingface.co/BabakScrapes/disco-clause-segmenter), which splits raw text into clauses first).
+This is the SE-classification component of the DiSCo pipeline (the precursor companion model is [`BabakScrapes/disco-clause-segmenter`](https://huggingface.co/BabakScrapes/disco-clause-segmenter), which splits raw text into clauses first).
+
+See the [demos](https://huggingface.co/spaces/BabakScrapes/Anecedotal_Discourse_Classifier_Demo).
 
 ## Label set
 
@@ -48,7 +50,7 @@ The model outputs 18 classes. The mapping to `(genericity, eventivity, boundedne
 | 16 | NONSENSE | NA | NA | NA |
 | 17 | OTHER | NA | NA | NA |
 
-Classes 14–17 carry no attribute decomposition and are excluded from attribute-share calculations in the DiSCo analyses.
+Classes 14–17 carry no attribute decomposition per the definitions of the linguistic attributes and are excluded from attribute-share calculations in the DiSCo analyses.
 
 ## Usage
 
@@ -63,12 +65,12 @@ clause = "my friend smoked marijuana daily"
 enc = tok(clause, return_tensors="pt", truncation=True, max_length=128)
 with torch.inference_mode():
     pred_id = model(**enc).logits.argmax(-1).item()
-print(model.config.id2label[pred_id])   # e.g. "GENERALIZING SENTENCE (DYNAMIC)" -> (specific, dynamic, habitual)
+print(model.config.id2label[pred_id])   
 ```
 
 ## Training data
 
-Fine-tuned on the **DiSCo corpus** clause annotations — opinionated, mixed-register English text (news from four outlets across the political spectrum, Reddit, and AI-generated text) primarily on the topic of marijuana legalization. This domain is deliberately harder and more varied than the Wikipedia-dominated SitEnt corpus on which prior SE classifiers were trained.
+Pre-trained on the SitEnt corpus, then fine-tuned on the **DiSCo corpus** clause annotations: opinionated, mixed-register English text (news from four outlets across the political spectrum, Reddit, and AI-generated text) primarily on the topic of marijuana legalization. This domain is deliberately harder and more varied than the Wikipedia-dominated SitEnt corpus on which prior SE classifiers were trained.
 
 ## Performance
 
@@ -83,14 +85,13 @@ Training-time validation performance (seed 42, random 10% validation):
 
 The 18-way macro-F1 is depressed by severe class imbalance (the most frequent labels are ~10× more common than the rarest); the per-attribute metrics better reflect what downstream analyses consume.
 
-**Note on evaluation.** These are *training-time validation* numbers. The full DiSCo corpus predates the model's training and the original train/validation indices are not retrievable, so an independent within-corpus hold-out is not available. The released weights and corpus let you re-train and re-evaluate under any split you prefer.
+**Note on evaluation.** These are *training-time validation* numbers. The released weights and corpus let you re-train and re-evaluate under any split you prefer.
 
 ## Limitations
 
-- English only; tuned for opinionated, mixed-register text on controversial policy topics. Performance on other genres may differ.
-- Expects a single clause as input; pair with the DiSCo segmenter for raw text.
-- Not a content/topic model — the features are formal and content-agnostic by design.
+- English only
+- pre-trained on largely encyclopedic texts then tuned for opinionated, mixed-register text on one controversial policy topic. Performance on other genres may differ. However, as the features are formal and content-agnostic by design, good cross-genre generalization is expected.
 
 ## Citation
 
-Hemmatian, B. (2022). *[Dissertation title]*. Brown University. And the DiSCo corpus paper (forthcoming, *Behavior Research Methods*). See the [DiSCo demos](https://huggingface.co/spaces/BabakScrapes/Anecedotal_Discourse_Classifier_Demo).
+Hemmatian, B. (2022). *Taking the High Road: A Big Data Investigation of Natural Discourse in the Emerging U.S. Consensus about Marijuana Legalization*. Brown University. And the DiSCo corpus paper (forthcoming, *Behavior Research Methods*). See the [DiSCo demos](https://huggingface.co/spaces/BabakScrapes/Anecedotal_Discourse_Classifier_Demo).