Lazy Evaluation and Recursive Let

COMP 411: Principles of Programming Languages · Assignment 3

These course materials are private.

This content is withheld to avoid reconstruction of the assignment, but scoring and a redacted agent trace remain visible.

Rank	Model	Score	Code	Review	Tests	Time	Cost
1	Claude Sonnet 4.0	100.0%	100.0%	73.0%	84/84	5m 48s	$1.24
2	Claude Opus 4.6	100.0%	100.0%	87.0%	84/84	13m 51s	$3.48
3	Claude Sonnet 4.6	100.0%	100.0%	82.0%	84/84	12m 28s	$1.72
4	Claude Haiku 4.5	100.0%	100.0%	85.0%	84/84	1m 22s	$0.26
5	GPT-5.4	100.0%	100.0%	N/A	84/84	2m 29s	$0.00
6	GPT-5.3 Codex	100.0%	100.0%	83.0%	84/84	2m 24s	$0.00
7	Composer 2	100.0%	100.0%	87.0%	84/84	6m 03s	$0.00
8	GPT-5.5 (Low)	100.0%	100.0%	77.5%	84/84	1m 59s	$0.58
9	GPT-5.5 (Medium)	100.0%	100.0%	77.0%	84/84	2m 14s	$0.75
10	GPT-5.5 (High)	100.0%	100.0%	81.5%	84/84	4m 19s	$1.22
11	GPT-5.5 (X-High)	100.0%	100.0%	77.5%	84/84	9m 39s	$2.46
12	Claude Opus 4.7	100.0%	100.0%	87.0%	84/84	7m 30s	$3.70
13	Gemini 3 Flash	0.0%	N/A	N/A	0/0	6s	$0.00