23 Mar 2026
Job Information
Organisation/Company: INESC ID
Research Field: Engineering » Computer engineering
Researcher Profile: First Stage Researcher (R1)
Positions: Master Positions
Application Deadline: 30 Sep 2026 - 23:59 (Europe/Lisbon)
Country: Portugal
Type of Contract: Not Applicable
Job Status: Not Applicable
Is the job funded through the EU Research Framework Programme?: Not funded by a EU programme
Reference Number: BI|2026/881 - Projet SALVE – refª 2024.14936.PEX
Is the Job related to staff position within a Research Infrastructure?: No
Offer Description
Public notice for one research grant Refª BI|2026/881
Projet SALVE – refª 2024.14936.PEX
funded by Fundação para a Ciência e a Tecnologia
Application submission from 23-Mar to 30-Sept 2026
WHO WE ARE
INESC-ID (www.inesc-id.pt), “Instituto de Engenharia de Sistemas e Computadores: Investigação e Desenvolvimento em Lisboa” is a Research and Development and Innovation Organization (R&D+i) in the fields of Computer Science and Electrical and Computer Engineering.
INESC-ID’s research impact is focused on four Thematic Lines
- Energy transition
- Life and health technology
- Security and privacy
- Societal digital transformation
INESC-ID promotes cooperation between academia and industry by addressing research on daily life issues, such as healthcare, space, mobility, human language technologies, agri-food, industry 4.0, and smart grids. This high level of knowledge transfer is achieved through both competitive research projects and direct contracted research. Therefore, public and private entities have access to a pool of knowledge, resources and services provided through the unique competencies available at the institution.
INESC ID has a strong international academic environment and inclusive work environment.
ONE (1) research grant for students with MSc degree with reference number BI|2026/881 under the scope of the Projet SALVE: Securing Artificial Language Models Against Vulnerability Encoding (2024.14936.PEX), funded by Fundação para a Ciência e a Tecnologia, is available under the following conditions:
OBJECTIVES | FUNCTIONS
This task aims to enhance the ability of Large Language Models (LLMs) to distinguish secure from insecure JavaScript code using contrastive learning with a tailored security-aware loss function. The student will fine-tune selected models using secure-insecure code pairs derived from Tasks 1 and 2 and evaluate improvements in classification stability and security-aware code generation.
The work plan includes:
- (Month 1) Implement contrastive learning fine-tuning using a tailored Multiple Negatives Ranking Loss (MNRL) formulation.
- (Month 2) Design and integrate a security penalty term to balance false positives and false negatives.
- (Month 3) Analyze embedding-space separation using cosine similarity and alternative visualization techniques.
- (Month 4) Evaluate improvements in classification metrics (accuracy, precision, recall, F1, FNR, FPR).
- (Month 4) Compare fine-tuned models against baseline models without contrastive learning.
- (Month 5) Assess secure-by-default code generation using static analysis tools (e.g., Semgrep, CodeQL), measuring vulnerabilities per 100 lines of generated code.
- (Month 6) Ensure reproducibility and open-source release of training and evaluation pipelines.
The selected candidate will be integrated into a research team with established expertise in software security, program analysis, and AI-driven code intelligence, with a track record of collaboration with leading technology companies and publications in top-tier international conferences and journals.
BINDING LEGISLATION
Law 40/2004 of 18th of August (Scientific Research Fellow Status) in its current wording.
https://diariodarepublica.pt/dr/legislacao-consolidada/lei/2004-58216179
Regulations for Research Grants of the Foundation for Science and Technology in force
https://diariodarepublica.pt/dr/detalhe/regulamento/950-2019-127238533
INESC-ID Lisboa Grant Regulations in force
https://www.inesc-id.pt/scholarship-regulations/
The fellowship contract is awarded on an exclusive dedication basis – art. 5 of Scientific Research Fellow Status and art. 16 of Regulations for Research Grants of the Foundation for Science and Technology.
DURATION
SIX (6) months, starting in November 2026, renewable subject to suitable performance within the period of the project, not exceeding the maximum period set by FCT for such grants.
RENEWALL
Renewable is subject to performance if the candidate is enrolled in a PhD program - art. 6º, n.4 c)
https://diariodarepublica.pt/dr/detalhe/regulamento/950-2019-127238533
https://files.dre.pt/2s/2019/12/241000000/0009100105.pdf
subject to suitable performance within the period of the project, not exceeding the maximum period set by FCT for such grants – 4 years (included contract renewals)
Renewable is subject to performance if the candidate is enrolled in a Non-degree programme – art. 6º, n.4 a)
https://diariodarepublica.pt/dr/detalhe/regulamento/950-2019-127238533
https://files.dre.pt/2s/2019/12/241000000/0009100105.pdf
subject to suitable performance within the period of the project, not exceeding the maximum period set by FCT for such grants – 1 year (included contract renewals)
MONTHLY AMOUNT
The monthly amount of the grant is 1309,64€ according to the “Regulations for Research Grants of the Foundation for Science and Technology” in force.
The payment will be made through a monthly bank transfer to a portuguese bank account, held by the grantee.
https://www.fct.pt/wp-content/uploads/2025/02/Tabela_valores_SMM_2025.pdf
SCIENTIFIC SUPERVISION
The activity will be supervised by:
- Rui Abreu, Full Professor at FEUP and Researcher at INESC ID.
- Sofia Reis, Invited Professor at FEUP
WORKPLACE
- INESC ID - Headquarters Rua Alves Redol, 9, 1000-029 Lisboa
- INESC-ID - Taguspark Avenida Professor Cavaco Silva, Edifício IST 2744-016 Porto Salvo
- INESC-ID Energia IST - Departamento de Engenharia Electrotécnica e de Computadores, Área de Energia, Av. Rovisco Pais 1049-001 Lisboa
ELIGIBILITY CONDITIONS
Any national, foreign and stateless candidate(s) holding a MSc degree in Computer Science, Software Engineering, Artificial Intelligence, or Cybersecurity, or closely related fields, and with a suitable scientific and professional curriculum may apply.
The candidate(s) must be enrolled in a
- PhD program - art. 6º, n.4 b)
https://diariodarepublica.pt/dr/detalhe/regulamento/950-2019-127238533 - Non-degree programme – art. 6º, n.4 a)
https://diariodarepublica.pt/dr/detalhe/regulamento/950-2019-127238533
The applicants are required to have proficiency in Portuguese and English, both written and spoken, as well as being enthusiastic, proactive, collaborative and able to work individually and as a team.
Preferential factors:
Preference will be given to candidates who demonstrate one or more of the following:
- Familiarity with software security and vulnerability detection, particularly in JavaScript or web security contexts (e.g., XSS, injection classes), and/or experience using static analysis tools (e.g., Semgrep, CodeQL).
- Experience with empirical evaluation of ML models (e.g., precision/recall/F1, robustness/stability analysis, and error analysis).
- Strong experimental and reproducibility practices.
- Hands-on experience training or fine-tuning neural models (e.g., Hugging Face ecosystem, PEFT/LoRA, custom training loops).
- Experience with representation learning and contrastive learning, including familiarity with metric learning objectives (e.g., triplet loss, InfoNCE, MNRL) and similarity-based evaluation (e.g., cosine similarity).
- Experience designing or modifying loss functions, especially for classification robustness, imbalance-aware learning, or security-relevant objectives (e.g., penalizing false negatives).
EVALUATION CRITERIA
The selection will be based on the following criteria:
- Academic background and relevant professional experience – 40%
Evaluation of academic performance and/or relevant professional experience in machine learning, AI, software engineering, or security-related domains. - Research track record and scientific contributions – 15%
Evaluation of publications, thesis work, or prior research contributions relevant to machine learning, representation learning, or software security.
(For MSc candidates, thesis or strong research projects will be considered.) - Technical expertise in machine learning and model fine-tuning – 10%
Demonstrated experience with neural network training, loss function design, embedding-based models and fine-tuning large language models. - Experience with empirical evaluation and experimental reproducibility – 10%
Demonstrated ability to design controlled experiments, evaluate models rigorously, analyze classification metrics and build reproducible ML pipelines. - Background in software security or vulnerability detection – 5%
Relevant coursework, projects, or professional experience. - Motivation letter and alignment with project objectives – 20%
Assessment of: Understanding of contrastive learning and security-aware model training; Intellectual alignment with the project’s objectives; Clarity of research interest and commitment to contributing to model improvement.
The members of the evaluation panel assign a classification to each of the candidates on a scale of 0 to 100 points, ranking the candidates according to their classification consisting on the sum of the partial classifications assigned in each evaluation criterion, and considering the weighting factor given to each parameter. In this process abstentions are not allowed.
In the event of a tie among candidates with the same highest evaluation score, the Evaluation Panel reserves the right to conduct interviews to have a more objective assessment of specific criteria.
If an interview is conducted, the punctuation proposed by the evaluation panel is obtained by assigning a weighting factor of 80% to the CV, and a weighting factor of 20% to the interview.
Additionally, the evaluation panel retains the discretion to withhold the scholarship if the candidates fail to fulfil the stipulated requirements.
The evaluation panel may also decide not to assign the scholarship, if none of candidates meets the required conditions.
If the requirements mentioned above are not fulfilled, the evaluation panel may choose not to select a candidate.
Candidates with an average score below 70, are not recommended for hiring
Minutes of the evaluation panel meeting are drawn up with a summary of what has taken place in the meeting, as well as the votes emitted by the jury, being available to the candidates whenever requested.
EVALUATION PANEL
Jury
- President: Miguel Pupo Correia – Researcher / Full Professor – INESC-ID / Tecnico, ULisboa
- Member: Nuno Santos – Researcher / Associate Professor – INESC-ID / Tecnico, ULisboa
- Member: João F. Ferreira – Researcher / Associate Professor – INESC-ID / FEUP
- Substitute Member: José Fragoso Santos – Researcher / Assistant Professor – INESC-ID / Tecnico, ULisboa
- Substitute Member: Rui Abreu – Researcher / Full Professor – INESC-ID / FEUP
APPLICATION
The application and supporting documents must be submitted via email to bolsas@inesc-id.pt.
The application email should clearly state the reference of the grant on the subject.
Refª BI|2026/881 - Projet SALVE – refª 2024.14936.PEX
Applications without the reference of the grant will not be considered.
The candidate should clearly state his / her consent so that the communications and notifications in the context of this call can take place to the electronic address bolsas@inesc-id.pt.
The application and supporting documents may be submitted in Portuguese or English, although the President of the evaluation panel when a member of the evaluation panel is a non-native Portuguese speaker, may request that, within a reasonable time, the candidate translates into English a document previously presented in Portuguese.
Candidates who submit their application incorrectly or who do not prove the requirements required in this public notice are excluded from admission. In case of doubt, the evaluation panel may demand any candidate to present documents proving those statements.
False statements by the candidates will be punished according to the law.
SUPPORTING DOCUMENTS
The supporting documents listed below must be presented in pdf form. Supporting documents presented in another format will not be considered by the evaluation panel.
- Motivation letter
Addressed to the Evaluation Panel, including full name, and contact address, including email address and telephone. - Detailed curriculum vitae;
- Single copy of official academic degree certificate required in the ELIGIBILITY CONDITIONS
In the application submission, the candidates may replace the copy of official academic degree certificate by a declaration of honour, stating that they have the required academic degree. It is mandatory for the approval of the fellowship contract that the selected candidate presents a single copy of the official academic degree certificate, required in education level.
In the application submission, the candidates from foreign education institutions may replace the copy of the official academic degree certificate by declaration of honour, stating that they have the required academic degree. It is mandatory for the approval of the fellowship contract that the selected candidate presents a single copy of the official diploma required in education level.
- Detailed list of grades
- Proof of enrolment required in the ELIGIBILITY CONDITIONS at grant starting date;
In the application submission, the candidates may replace the proof of enrolment by a declaration of honour, stating that they are/will be enrolled in the program required ELIGIBILITY CONDITIONS. It is mandatory for the approval of the fellowship, that the selected candidate present the official proof of enrolment required in the ELIGIBILITY CONDITIONS.
NOTIFICATION, PRIOR HEARING, COMPLAIN AND APPEAL DEADLINES
The admitted and excluded candidates will be notified by email of the final ranking list.
After being notified, all candidates have 10 working days to submit a formal rebuttal. After that period, the evaluation panel notifies the candidates of the final report.
Excluded candidates may complain about the evaluation panel final report for 15 working days after notification or appeal to the evaluation panel decision to the INESC ID Board of Directors for 30 working days, after notification.
Non-discrimination and equal access policy
INESC-ID actively promotes a policy of non-discrimination and equal access, so that no candidate can be privileged, benefited, disadvantaged or deprived of any right or exemption from any duty owing, in particular, to ancestry, age, sex, sexual orientation, marital status, family status, economic situation, education, social origin or condition, genetic heritage, reduced working capacity, disability, chronic illness, nationality, ethnic origin or race, territory of origin, language, religion, political or ideological beliefs and trade union membership.
Under the terms of D.L. No. 29/2001, of February 3, a disabled candidate has a preference when in equal classification, which prevails over any other legal preference. Candidates must declare on the application form their respective degree of disability, the type of disability and the means of communication / expression to be used in the selection process, under the terms of the aforementioned diploma.
Requirements
Research Field: Engineering » Computer engineering
Education Level: Master Degree or equivalent
Work Location(s)
Number of offers available: 1
Company/Institute: INESC ID
Country: Portugal
State/Province: Lisbon
City: Lisbon
Postal Code: 1000-029
Street: Rua Alves Redol, 9
Contact
State/Province: Lisboa
City: Lisboa
Website: http://www.inesc-id.pt
Street: Rua Alves Redol, 9
Postal Code: 1000-029
E-Mail: bolsas@inesc-id.pt
This is a Preview Listing…
You must sign in to see the full job description, and to apply.
Manage / Upgrade this job to a Full Job Listing.
Find Your Best Opportunity
Tell them AcademicJobs.com sent you!