BPSC 69th CCE Prelims General Studies · Question 6 of 144

Correct — D, Generative Pre-Trained Transformer. ChatGPT is OpenAI's conversational front end to its GPT family of large language models, and each of the three words in the expansion names a separate design decision, which is why the full form is worth knowing rather than merely recognising. Generative means the model produces new text instead of only classifying or scoring text that already exists: it is trained to predict the next token — roughly a word-piece — given everything before it, and writing an answer is simply that prediction run over and over. Pre-Trained names the two-stage recipe OpenAI set out in its 2018 paper 'Improving Language Understanding by Generative Pre-Training' (Radford, Narasimhan, Salimans and Sutskever): an expensive unsupervised first stage in which the model reads an enormous unlabelled corpus and absorbs general language structure, then a far cheaper second stage adapting it to a task — for ChatGPT, supervised fine-tuning followed by reinforcement learning from human feedback (RLHF), which is what teaches it to follow an instruction and to decline an unsafe request. Transformer is the neural-network architecture introduced in the Google paper 'Attention Is All You Need' (Vaswani and co-authors, 2017); its self-attention mechanism lets every token be weighed directly against every other token in parallel, replacing the step-by-step recurrence of earlier models and making training at internet scale practical on GPUs. The scale-up was rapid — GPT-1 had 117 million parameters in 2018, GPT-2 1.5 billion in 2019 and GPT-3 175 billion in 2020. ChatGPT itself was released as a free research preview on 30 November 2022 running on GPT-3.5, GPT-4 followed on 14 March 2023, and the product was estimated to have crossed 100 million monthly users by January 2023 — which is exactly why the acronym was 'recently seen in the news' when this paper was set in September 2023.

• (a)Glutamic Pyruvic Transaminase — A genuine GPT, and the most tempting wrong answer for anyone with a biology background: glutamic-pyruvic transaminase is the liver enzyme now more usually called alanine aminotransferase (ALT), and Indian pathology reports still print it as SGPT in a liver function test, where a raised value points to hepatocellular injury such as viral hepatitis or drug damage. It is decades older than the AI sense but has nothing to do with language models, and the stem's qualifier 'in ChatGPT' rules it out.
• (b)GUID Partition Table — Also a real GPT, and the closest trap inside computing itself. The GUID Partition Table is the disk-partitioning layout defined by the UEFI specification, brought in to replace the legacy Master Boot Record scheme, which could address only about 2 TB of disk and hold four primary partitions; GPT uses 64-bit addressing and stores a backup copy of the table at the end of the disk. It is a storage format, not a model architecture, and nothing about it was in the news in 2023.
• (c)Grooved Pegboard Test — A third real expansion, this one from clinical neuropsychology: the Grooved Pegboard Test times how quickly a subject can rotate keyed pegs and drop them into randomly oriented slots, and is used to measure fine motor speed and eye-hand coordination, often to detect impairment on one side of the body. Its presence here shows the examiner's design — every option is a true GPT, so elimination of nonsense will not work and only the correct sense of the acronym does.

A large language model (LLM) is a neural network trained on a very large body of text to predict the next token given everything before it; ChatGPT is such a model wrapped in a chat interface. Its architecture — the transformer — works by self-attention: for every token the network computes how strongly it should be influenced by every other token in the input, so context is captured directly rather than passed hand to hand along a chain, and the whole sequence can be processed in parallel. That parallelism is why transformers scaled where the earlier recurrent networks (RNNs and LSTMs) stalled. Training runs in two phases, a broad unsupervised pre-training phase and a small alignment phase using human-ranked responses. Models built this way are called foundation models, because one pre-trained base is adapted to many downstream tasks, and generative AI, because the output is new content rather than a label. The same architecture now underlies image, speech and protein-structure models, not text alone.

This is an acronym-expansion item and all four options are real expansions of GPT — that is the whole design. The examiner is not checking whether you can discard nonsense; he is checking whether you know which sense of the acronym belongs to ChatGPT. Two discriminators settle it. First, read the qualifier: 'in ChatGPT' and 'recently seen in the news' both point at the computing sense, which removes (a), a clinical enzyme familiar to any pathology lab since long before 2023. Second, separate storage from intelligence within computing itself — (b) GUID Partition Table is a way of laying out partitions on a disk, an unglamorous standard that no newspaper covered. What is left is (d), and it can also be reached positively rather than by elimination: 'transformer' is the only word among the four options that names a neural-network architecture, and 'generative' is precisely the property that made ChatGPT famous, because it writes rather than merely classifies. Candidates lose this question by pattern-matching on the biology option, since SGPT is drilled into everyone through the liver function test.

• GPT expands to Generative Pre-Trained Transformer; the phrase first appears in OpenAI's 2018 paper 'Improving Language Understanding by Generative Pre-Training' by Radford, Narasimhan, Salimans and Sutskever.
• The transformer architecture was introduced in 'Attention Is All You Need' (Vaswani and co-authors, Google, 2017); its core idea, self-attention, removed the need for recurrence and made parallel training on GPUs possible.
• Parameter growth of the series: GPT-1 117 million (2018), GPT-2 1.5 billion (2019), GPT-3 175 billion (2020). OpenAI has never published a parameter count for GPT-4, which was released on 14 March 2023.
• ChatGPT was launched as a free research preview on 30 November 2022 running on GPT-3.5, and was estimated to have reached about 100 million monthly users by January 2023 — the fastest uptake recorded by any consumer application to that point.
• India's policy response came soon after: the Union Cabinet approved over Rs 10,300 crore for the IndiaAI Mission on 7 March 2024, spread over five years across seven pillars, the compute pillar alone targeting the deployment of more than 10,000 GPUs through public-private partnership.

• All four options are genuine expansions of GPT — the item is decided by the qualifier 'in ChatGPT', not by spotting a fake
• Treating GPT and LLM as synonyms: GPT names one family of transformer-based models built by OpenAI, while LLM is the general category and transformer is the architecture underneath both
• Confusing the product with the model — ChatGPT is the chat interface; GPT-3.5, GPT-4 and later releases are the models it runs on

BPSC asks technology as one-line 'in the news' recall — the full form, the launch year, the parent company — and then stacks the options with other genuine expansions of the same letters so that half-recognition fails. UPSC almost never asks for a full form; it asks what a technology can and cannot presently do, as in 2020's 'With the present state of development, Artificial Intelligence can effectively do which of the following?', or buries one AI statement among statements on unrelated technologies and makes you judge each.

• practice — not a real PYQ

  The IndiaAI Mission, approved by the Union Cabinet in March 2024, is implemented under which Union Ministry?

  • (a)Ministry of Science and Technology
  • (b)Ministry of Electronics and Information Technology
  • (c)Ministry of Education
  • (d)Ministry of Commerce and Industry

  Answer(b) Ministry of Electronics and Information Technology — MeitY runs the IndiaAI Mission, approved on 7 March 2024 with an outlay of over Rs 10,300 crore for five years.

• practice — not a real PYQ

  In a liver function test, the enzyme reported as SGPT is also known as:

  • (a)Alanine aminotransferase (ALT)
  • (b)Aspartate aminotransferase (AST)
  • (c)Alkaline phosphatase (ALP)
  • (d)Gamma-glutamyl transferase (GGT)

  Answer(a) Alanine aminotransferase (ALT) — serum glutamic-pyruvic transaminase and ALT are the same enzyme; the companion marker SGOT corresponds to AST.