IAESTE Internships

Location

India

Internship type

ON-SITE

Reference number

IN-2026-B1116-KU

Students Requirements

General discipline

Computer Science / Informatics

Completed Years of Study

3

Fields of Study

General
Computer and Information Research

Languages

English Excellent (C1, C2)

Required Knowledge and Experience

-

Other Requirements

The intern should have knowledge related to Python, deep learning frameworks (PyTorch/TensorFlow), Large Language Models, reinforcement learning basics, and neural architecture search concepts.

Work Details

Duration

20 - 20 Weeks

Within These Dates

04.05.2026 - 05.10.2026

Holidays

NONE

Work Environment

-

Gross pay

10000 INR / month

Working Hours

40.0 per week / 8.0 per day

Living Lodging

Type of Accommoditation

IAESTE- LC KARUNYA

Cost of lodging

5000 INR / month

Cost of living

8000 INR / month

Work Offered

Additional Info

This offer is from the Department of Computer Science, and the participant’s field of research would be autonomous neural architecture discovery through LLM-driven proposal engines, rapid evaluation pipelines, and reinforcement learning integration.

Work description

Autonomous Research Agent for Neural Architecture DiscoveryOverview:This project aims to automate the design of advanced neural networks, a process that typically requires years of human-led experimentation. The goal is to build an autonomous research agent capable of proposing, evaluating, and refining architectures using Large Language Models (LLMs), rapid evaluation pipelines, and reinforcement learning. Acting as an AI researcher, it will rediscover canonical designs while uncovering innovative and efficient architectures, advancing the frontiers of AutoML and scientific automation.Objectives:1) Define and structure the search space for architecture exploration.2) Implement an LLM-based proposal engine for generating candidate models.3) Develop a reinforcement learning pipeline for iterative refinement.Outcomes:1) Rediscovery of established neural architectures.2) Discovery of novel, efficient, high-performing models.3) Practical insights into combining LLMs, RL, and AutoML.Responsibilities:1) Build and refine the architecture search space.2) Implement LLM-driven proposal and evaluation systems.3) Integrate modules into a cohesive RL-based framework.

Deadline

26.04.2026