StanfordSCALE/assertion_contains_math_computation

Assertion: contains math computation

This classifier was trained for EduBehaviors: Assertion-based schemas for auditable dialogue coding and is usable through the Python package EduBehaviors-kit. This classifier was trained on an LLM-annotated subset of teacher utterances from the TalkMoves Dataset. See the Datasets section below for more information.

---

Training Details

Datasets

Dataset	Split	Size
StanfordSCALE/assertions_llm_annotated_talkmoves	train	13,184 (53.3%)
StanfordSCALE/assertions_llm_annotated_talkmoves	dev	3,295 (13.3%)
StanfordSCALE/assertions_llm_annotated_talkmoves	test	8,236 (33.3%)

This model's columns are assertion_contains_math_computation and split_contains_math_computation.

Base rate (share of rows labeled as True): 3.7% overall — 3.7% train, 3.9% dev, 3.7% test.

Labels and annotation

Labels were generated with LLM annotators. Krippendorff's alpha for this assertion is 0.564.

Hyperparameters

Parameter	Value
Base model (body)	sentence-transformers/paraphrase-mpnet-base-v2
Head	LogisticRegression
Body learning rate	2e-05
Head learning rate	0.01
Batch size	16 (contrastive phase) / 32 (head)
Epochs	10
Max steps	5000 (contrastive phase)
Eval max steps	100
Seed	20260904
Mixed precision	enabled on GPU

---

Evaluation

Results

Split	n	Base rate	Precision	Recall	F1 (positive class)	ROC-AUC	Average precision
dev	3,295	3.9%	0.669	0.654	0.661	0.915	0.690
test	8,236	3.7%	0.656	0.565	0.607	0.908	0.642

Limitations

• Labels come from LLM annotators, not human coders. Agreement between annotators with Krippendorff's Alpha is 0.564.
• Trained on teacher utterances only. Behaviour on student speech is untested.

---

How to Use

Message Structure

The model was trained on text built as:

{utterance}

The utterance is passed through as-is.

Running instructions

pip install setfit

from setfit import SetFitModel

model = SetFitModel.from_pretrained("StanfordSCALE/assertion_contains_math_computation")

text = 'Makes minus 54 X'
model.predict([text])        # -> array([1]) when the assertion holds
model.predict_proba([text])  # -> [[P(no), P(yes)]]

Citation

@misc{assertion_contains_math_computation,
  author = {Stanford SCALE Initiative},
  title  = {Assertion classifier: contains math computation},
  year   = {2026},
  url    = {https://huggingface.co/StanfordSCALE/assertion_contains_math_computation}
}