layout: page title: Assessment permalink: /assessment/ —

Overview

This module uses 8 assessment components totalling 100%.

Components

ID Title Type Mode Weight Due Week Delivery
Q1 In-Class Quiz 1: Probability and Entropy Foundations in-class-quiz summative 10% 2 in-class
W1 Worksheet 1: Socratic Dialogue on Boltzmann and Entropy mini-project summative 15% 2 take-home
Q2 In-Class Quiz 2: MaxEnt, Exponential Family, Landauer in-class-quiz summative 10% 5 in-class
W2 Worksheet 2: Maxwell’s Demon, MaxEnt, and the Exponential Family mini-project summative 15% 5 take-home
Q3 In-Class Quiz 3: Fisher Metric, Natural Gradient, Dually Flat Geometry in-class-quiz summative 10% 7 in-class
W3 Worksheet 3: Information Geometry and Thermodynamic Length mini-project summative 15% 7 take-home
Q4 In-Class Quiz 4: Multi-Information, I+H=C, Von Neumann Entropy, Limits on Intelligence in-class-quiz summative 10% 8 in-class
W4 Worksheet 4: Multi-Information, Von Neumann Entropy, and Limits on Intelligence mini-project summative 15% 8 take-home

Component Details

Q1: In-Class Quiz 1: Probability and Entropy Foundations

Type: in-class-quiz | Weight: 10% | Due week: 2 | Delivery: in-class Estimated workload: 0.2h AI robustness: Invigilated in-class test on Moodle. Random draw from a large bank reduces repeatability. No external resources.

W1: Worksheet 1: Socratic Dialogue on Boltzmann and Entropy

Type: mini-project | Weight: 15% | Due week: 2 | Delivery: take-home Estimated workload: 3.0h AI robustness: Students must pose curiosity and skepticism prompts as Socrates; authenticity is verified by the paired in-class Quiz 1 (Week 2), which tests probability/entropy foundations and Week 1 seeds under invigilated conditions.

Q2: In-Class Quiz 2: MaxEnt, Exponential Family, Landauer

Type: in-class-quiz | Weight: 10% | Due week: 5 | Delivery: in-class Estimated workload: 0.2h AI robustness: Invigilated in-class test on Moodle. Random draw from a large bank. No external resources.

W2: Worksheet 2: Maxwell’s Demon, MaxEnt, and the Exponential Family

Type: mini-project | Weight: 15% | Due week: 5 | Delivery: take-home Estimated workload: 3.0h AI robustness: LLMs are permitted. Authenticity verified by in-class Quiz 2 (Week 5), which applies MaxEnt and Landauer to a new problem under invigilated conditions.

Q3: In-Class Quiz 3: Fisher Metric, Natural Gradient, Dually Flat Geometry

Type: in-class-quiz | Weight: 10% | Due week: 7 | Delivery: in-class Estimated workload: 0.2h AI robustness: Invigilated in-class test on Moodle. Random draw from a large bank. No external resources.

W3: Worksheet 3: Information Geometry and Thermodynamic Length

Type: mini-project | Weight: 15% | Due week: 7 | Delivery: take-home Estimated workload: 3.0h AI robustness: LLMs permitted. Authenticity verified by in-class Quiz 3 (Week 7), applying Fisher metric and natural gradient to a new parameterised family under invigilated conditions.

Q4: In-Class Quiz 4: Multi-Information, I+H=C, Von Neumann Entropy, Limits on Intelligence

Type: in-class-quiz | Weight: 10% | Due week: 8 | Delivery: in-class Estimated workload: 0.2h AI robustness: Invigilated in-class test on Moodle. Random draw from a large bank. No external resources.

W4: Worksheet 4: Multi-Information, Von Neumann Entropy, and Limits on Intelligence

Type: mini-project | Weight: 15% | Due week: 8 | Delivery: take-home Estimated workload: 3.0h AI robustness: LLMs permitted. Authenticity verified by in-class Quiz 4 (Week 8), applying multi-information and the perpetual motion analogy to a new example under invigilated conditions.