Unleashing the Power of Generative AI Models ICT5358 T1 Assignment Help
AssessmentBrief:ICT5358UnleashingthePowerofGenerativeAIModels Trimester 1, 2025
AssessmentOverview
|
AssessmentTask |
Type |
Weight ing |
Length |
Due |
ULOs Assessed |
|
Assessment1:AppliedProject Reporttooutlinetheretailand wholesale application of large language models. |
Individual
|
20% |
1000 words |
Week 4 |
ULO1 ULO4 |
|
Assessment2:Laboratory Practicum |
Individual
|
30% |
1hour,1500 words + |
Weeks 6, 8, 9 |
ULO1 ULO2 |
|
Regularinvigilatedpracticaltests, implementingpromptengineering experimentsanddesigningAPIuse cases. |
Invigilated
|
|
Design (equiv. 1500 words) |
|
ULO3 ULO5 |
|
Assessment3:AppliedGenerative AI Model Reportoutliningthemethodology, |
Group
|
25% |
1000 words+ Design |
Week 7 |
ULO1 ULO2 ULO5 |
|
observations,resultsand |
|
|
(equiv. |
|
|
|
evaluationsofpromptengineering experiments. |
|
|
1000 words) |
|
|
|
Assessment4:Case Study Designdocumentoutlininga |
Individual
|
25% |
1000 words+ |
Week 12 |
ULO1 ULO2 |
|
proposedapplicationofgenerative |
|
|
Design |
|
ULO3 |
|
AIinabusinessusecase. |
|
|
(equiv. 1000 words) |
|
ULO4 ULO5 |
Noteforallassessmentstasks:
● Studentscangenerate/modify/createtextgeneratedbyAI.Theyarethenaskedto modify the text according to the brief of the assignment.
● Duringthepreparationandwritingofanassignment,studentsuseAItools,butmay not include any AI-generated material in their final report.
● AItoolsareusedbystudentsinresearchingtopicsandpreparingassignments,but all AI-generated content must be acknowledged in the final report as follows:
Assessment2:LaboratoryPracticum
|
Due date: |
Weeks6,8,9 |
|
Group/individual: |
Individual |
|
Wordcount/Time provided: |
1500words+Design(equiv.1500words) (1hour) |
|
Weighting: |
3×10% = 30% |
|
UnitLearning Outcomes: |
ULO1, ULO2, ULO3, and ULO5 |
Assessment2 Detail
During one hour of theweekly seminarin Weeks6, 8 and 9, you willbe providedwithsome practicalusecasesforalargelanguagemodel.You willneedtoperformbasicandadvanced prompt engineering to achieve some practical results within the use case. You may also be required to explain how an API could be utilised to enhance the automation in the process.
Theseassessments areheldduringeach relevantweekly seminarand they are invigilated.
Assessment2MarkingCriteriaand Rubric
Eachassessmentwillbemarked outof100and willbeweighted10%ofthetotalunitmark, totalling a combined weight of 3 x 10% = 30%.
Assessment2MarkingCriteriaand Rubric
|
Marking Criteria |
Not satisfactory (0- 49%)ofthecriterion mark |
Satisfactory (50-64%)ofthe criterion mark |
Good (65-74%)ofthe criterion mark |
Very Good (75-84%)ofthe criterion mark |
Excellent (85-100%)ofthe criterion mark |
|
Prompt |
Showsnoevidence of |
Showssome |
Showssome |
Showsevidenceof |
Showsevidenceof |
|
Engineering(30 |
experimentationor |
evidenceof |
evidenceof |
experimentation, |
insightful |
|
marks) |
promptengineering. |
experimentation |
experimentation |
leadingtofinal |
experimentation, |
|
|
|
duringprompt |
duringprompt |
prompttemplate. |
andlearningfrom |
|
|
|
engineering,but |
engineering. |
|
experiment |
|
|
|
withlittleor no |
|
|
outcomes. |
|
|
|
relationshipto |
|
|
Integrates |
|
|
|
promptsused. |
|
|
experimentation |
|
|
|
|
|
|
intotheprocessof |
|
|
|
|
|
|
producingaprompt |
|
|
|
|
|
|
template. |
|
Prompt Elements(30 marks) |
Providesnoneof these: – aclearrole/persona statement in the prompt template, related to the use case. – clear and relevant inputparametersto guidethetone,style etc. of the output – provides clear instructions for the outputformat,related to the task – providesanexample output, and instructs LLM to this example Doesnotcorrectlyuse any system prompts, or shows no understanding of the role of system prompts. |
Provides at least oneofthesewhen relevant to a prompt template, but with some limitations in clarity: – a clear role/persona statement in the prompttemplate, relatedtotheuse case. – clear and relevant inputparametersto guide the tone,style etc. of the output – provides clear instructionsforthe output format, related to the task – provides an example output, andinstructsLLM to this example Makessomeuseof system prompts and user prompts when relevant. Offers limited explanationofthe roleofthesystem anduserprompts. |
Provides at least twoofthesewhen relevant to a prompt template, but with some possible room for improvement in clarity: – a clear role/persona statement in the prompttemplate, relatedtotheuse case. – clear and relevant inputparametersto guidethetone,style etc. of the output – provides clear instructionsforthe output format, related to the task – provides an example output,andinstructsLLMto this example Makes good use of systempromptsand user prompts when relevant. Shows goodunderstanding of the role of the system and user prompts. |
Provides at least threeofthesewhen relevant to a prompt template, but with some possible room for improvement in clarity in at least one: – a clear role/persona statement in the prompttemplate, relatedtotheuse case. – clear and relevant inputparametersto guide the tone,style etc. of the output – provides clear instructionsforthe output format, related to the task – provides an example output, andinstructsLLM to this example Makes great use of system prompts and user prompts when relevant, includingtheuseof at least one example. Shows thorough understanding of the role of the system and user prompts. |
Providesallfourof these when relevant to a prompt template: – a clear role/persona statement in the prompttemplate, relatedtotheuse case. – clear and relevant inputparametersto guide the tone,style etc. of the output – provides clear instructionsforthe output format, related to the task – provides an example output, andinstructsLLM to this example Makes clear and effective use of system promptsand user prompts when relevant, including the use of insightful few-shot examples. Clearly explains the role of thesystemanduser prompts. |
|
Automating GenerativeAI (30 marks) |
Doesnotcreatea helpful prompt template.
Makesnoeffortto utilise prompt chaining.
Offers no understandingofthe function or role of a LLM API in the workflow. |
Creates a prompt template that somewhat contributes to the taskwhenrequired, shows little understanding of the function of the template.
Createsatleastone prompt chain to carryoutpartofthe taskwhenrequired. Maynotexplainthe logic of this chain. Suggests some understanding of the function and roleofaLLMAPIin the workflow. |
Creates a functioning prompt template when required,andshows understanding of the function of the template.
Createsatleastone prompt chain to carryoutpartofthe taskwhenrequired, and shows some understanding of how this prompt chain relates to steps to guide the model to theoutput.
Explains thefunction and role of a LLM API in the workflowwithsome accuracy when required. Shows someunderstanding of how function calling works. |
Creates a well- structuredprompt template when required, and explains the template.
Creates a series of inputsandoutputs toguidethemodel to the final output when required. Converts these toaneffectivechainof prompts. Correctly explains the function and roleofa LLMAPI in theworkflowwhen required, including the nature and purposeoffunction calls. |
Creates a clear and well-structured prompt template whenrequired,and insightfullyexplains the template.
Createsaclearand logical series of inputsandoutputs toguidethemodel to the final output when required. Converts these toaneffectivechainof prompts, and explains the purpose of each prompt step. Clearly and insightfullyexplains the function and roleofa LLMAPI in theworkflowwhen required, including the nature and purposeoffunction calls and of hyperparameters such as temperature. |
|
Clarity and Presentation (10 marks) |
Answersareunclear, disorganised, and difficult to understand. |
Answers are somewhatclear, but lack organisation. |
Answers are mostly clearandorganised. |
Answersareclear and organised. |
Answers are clear andorganised,easy to understand, and engaging or insightful. |
Assessment3: AppliedGenerative AIModel
|
Due date: |
Week7 |
|
Group/individual: |
Group |
|
Wordcount/Time provided: |
1000words+Design(equiv.1000words) |
|
Weighting: |
25% |
|
UnitLearning Outcomes: |
ULO-1,ULO-2,ULO-5 |
Assessment3 Detail
The Alpha-Omega Corporation (AO) is a global company that has products and services across many different industries. AO is currently exploring the utility of generative AI models in various ways, including automating its own digital products and services (e.g. AO offers a paidservicetocreatewhitepapers, blogposts,andother digitalarticlesfor consumers),and providing marketing for those products.
You will be given a brief to updatetheSolutionspage of AO’swebsite, tomarket their products and services. You will submit a report that includes prompt engineering experiments, methodologies and outcomes, including:
● Theuseofroles,personas andotherparameters
● Specifyingrequirementsintheoutput
● Few-shotprompting
● Promptchaining
Youwilluse both languagemodelsandtext-to-image modelstogeneratetext andimages for the AO solutions page.
Assessments 3MarkingCriteriaandRubric
Theassessment willbemarkedoutof100andwillbeweighted25% ofthetotalunitmark.
|
Marking Criteria |
Not satisfactory (0- 49%)ofthecriterion mark |
Satisfactory (50-64%)ofthe criterion mark |
Good (65-74%)ofthe criterion mark |
Very Good (75-84%)ofthe criterionmark |
Excellent (85-100%)ofthe criterion mark |
|
Prompt Engineering(20 marks) |
Shows no methodology or no evidence of experimentation or promptengineering. |
Presents limited overall methodology. Shows some evidence of experimentation during prompt engineering,but with little or no relationship to prompts used. |
Presents some overall methodology. Shows some evidence of experimentation during prompt engineering. |
Presents a reasonableoverall methodology. Showsevidenceof experimentation, leading to final prompt template. |
Presents a clear methodological approach to addressing the problempresented. Shows evidence of insightful experimentation, and learning from experiment outcomes. Integrates experimentation into the process of producingaprompt template. |
|
Prompt Template(10 marks) |
Doesnotcreatea helpful prompt template. |
Creates a prompt template that somewhat contributes to the task, shows little understanding of thefunctionofthe template. |
Creates a functioningprompt template, and shows understanding of the function of the template. |
Creates a well- structuredprompt template, and explains the template. |
Createsaclearand well-structured prompt template, and insightfully explains the template. |
|
Prompt Elements(20 marks) |
Provides none of these,butwithsome possible improvements to clarityinatleastone: – aclearrole/persona statement in the prompt template, related to the use case. – clear and relevant inputparametersto guidethetone,style etc. of the output – provides clear instructions for the outputformat,related to the task – providesanexample output, and instructs LLM to this example |
Provides at least oneofthese,but with some limitations in clarity: – a clear role/persona statement in the prompttemplate, relatedtotheuse case. – clear and relevant inputparametersto guide the tone,style etc. of the output – provides clear instructionsforthe output format, related to the task – provides an example output,andinstructsLLMto this example |
Provides at least two of these, but withsomepossible improvements to clarity: – a clear role/persona statement in the prompttemplate, relatedtotheuse case. – clear and relevant inputparametersto guidethetone,style etc. of the output – provides clear instructionsforthe output format, related to the task – provides an example output,andinstructsLLMto this example |
Provides at least threeofthese, but withsomepossible improvements to clarity in at least one: – a clear role/persona statement in the prompttemplate, relatedtotheuse case. – clear and relevant inputparametersto guide the tone,style etc. of the output – provides clear instructionsforthe output format, related to the task – provides an example output,andinstructsLLMto thisexample |
Providesallfour: – a clear role/persona statement in the prompttemplate, relatedtotheuse case. – clear and relevant inputparametersto guide the tone,style etc. of the output – provides clear instructionsforthe output format, related to the task – provides an example output,andinstructsLLMto this example |
|
PromptChaining (20 marks) |
Makesnoeffortto utilise prompt chaining. |
Createsatleastone prompt chain to carryoutpartofthe task. May not explain the logic of this chain. |
Creates at least one prompt chain to carryoutpartofthe task, and shows someunderstanding of how this prompt chain relates to steps to guide the model to theoutput. |
Creates a series of inputs and outputs to guide the model to the final output. Converts these toaneffectivechainof prompts. |
Creates a clear and logical series of inputs and outputs to guide the model to the final output. Converts these toaneffectivechainof prompts, and explains the purpose of each promptstep. |
|
Evaluationof Results (15 marks) |
Does not offer any comparison of engineeredprompts to other prompts tried. |
Shares results of engineeredprompts and experimental results, but with littleunderstanding of the difference. |
Shares results of engineeredprompts and experimental results, with some understanding of the difference. |
Comparesresultsof engineeredprompts to experimental results. Draws conclusions about prompt features learnt. |
Insightfully comparesresultsof engineeredprompts to experimental results. Draws insightful conclusions about prompt features learnt. |
|
Clarity and Presentation (15 marks) |
Answersareunclear, disorganised, and difficult to understand. |
Answers are somewhatclear, but lack organisation. |
Answers are mostly clearandorganised. |
Answersareclear and organised. |
Answers are clear andorganised,easy to understand, and engaging or insightful. |
Assessment4:Case Study
|
Due date: |
Week12 |
|
Group/individual: |
Individual |
|
Wordcount/Time provided: |
1000words+Design(equiv.1000words) |
|
Weighting: |
25% |
|
UnitLearning Outcomes: |
ULO-1,ULO-2,ULO-3,ULO-4,ULO-5 |
Assessment4 Detail
The Alpha-Omega Corporation (AO) is a global company that has products and services across many different industries. AO is currently exploring the utility of generative AI models in various ways, including automating its own digital products and services (e.g. AO offers a paidservicetocreatewhitepapers, blogposts,andother digitalarticlesfor consumers),and providing marketing for those products.
You will create adesigndocument to helpAOto createan automated article writer tool. This tool allows the user to specify the type of article being created, including the styleand target audience, and specific information about the subject matter, and in response the tool will produce the article as a downloadable PDF for the user. In your design document, you’ll outline:
● An outline oftheworkflowof anautomated systemthatincludesoneormore generative AI models to power the generation of language and/or images
● Examples of promptsandprompttemplates thatcouldbeusedfor specifictasks performed by the tool, and how these integrate into the workflow
● Instructions, actionsandknowledgethatcould beintegratedintoacustomised generative AI model
● An evaluationoftheethicalissuesof building thetool,suchascopyrightand intellectual property rights
Assessments 4MarkingCriteriaandRubric
Theassessment willbemarkedoutof100andwillbeweighted25% ofthetotalunitmark.
|
Marking Criteria |
Not satisfactory (0- 49%)ofthecriterion mark |
Satisfactory (50-64%)ofthe criterion mark |
Good (65-74%)ofthe criterion mark |
Very Good (75-84%)ofthe criterionmark |
Excellent (85-100%)ofthe criterion mark |
|
GenerativeAI |
Unabletoidentify |
Abletoidentifythe |
Abletoidentifythe |
Abletoexplainhow |
Abletoclearlyand |
|
integrationand |
howaLLMwould |
basicsofhowaLLM |
essentialfeatures of |
aLLMwould |
insightfullyexplain |
|
workflow(20 |
integrateintotheuse |
wouldintegrate |
howaLLMwould |
integrateintothe |
howaLLMwould |
|
marks) |
case.Showsno |
intotheusecase. |
integrateintothe |
usecase, including |
integrateintothe |
|
|
understandingofthe |
Showspoororno |
usecase. Shows |
inputsandoutputs. |
usecase, including |
|
|
workflow. |
understandingof |
some |
Showsgood |
relevantand |
|
|
|
theworkflow. |
understandingof |
understandingof |
insightfulinputs |
|
|
|
|
theworkflow. |
theworkflow,with |
andoutputs.Shows |
|
Clarityand |
Answersareunclear, |
Answersare |
Answersaremostly |
Answersareclear |
Answersareclear |
|
Presentation |
disorganised,and |
somewhatclear, |
clearandorganised. |
and organised. |
andorganised, easy |
|
(10marks) |
difficultto |
butlack |
|
|
tounderstand, and |
|
|
understand. |
organisation. |
|
|
engagingor |
|
|
|
|
|
|
insightful. |
Leave A Comment