‘What just happened?': IIT Kharagpur student tests ChatGPT o3 on JEE Advanced mock test, taken aback by results
Recently, an IIT Kharagpur student tested ChatGPT o3 during her JEE Advanced 2025 mock test, and the results were shocking. In a blog post on software platform Heltar, Anushka Aashvi revealed that the model scored an astonishing 327 out of 360, a result that would have secured All India Rank 4 in the real exam.
Titled 'ChatGPT o3 Scores AIR 4 in JEE Advanced 2025. What Just Happened?', Aashvi shared that she went to great lengths to create a credible exam situation. The model was directed to 'act like a JEE aspirant,' solving each question separately with no internet access and no memory from previous answers. Every question was solved in a fresh chat session to prevent any form of carryover learning.
'We tested the ChatGPT o3 model (which was released on 16th April 2025) on the JEE Advanced 2025 question paper which was conducted on 18th May to ensure that the questions have as much newness for the AI model as possible,' Aashvi wrote.
Despite these constraints, ChatGPT o3 impressed at nearly every step. The platform helped her achieve perfect scores in Chemistry and Mathematics during the second half of the paper, and she lost only a few marks in Physics. The model showed a clear, step-by-step reasoning process, approaching multi-concept questions, advanced calculus problems, and even skeletal chemical diagrams.
'It easily solved lengthy algebra and calculus problems. The model performed remarkably well at combining concepts from multiple chapters to reach a correct solution. It was even able to interpret compounds correctly from their skeletal formulae and solve them correctly,' the student wrote in the blog.
However, ChatGPT o3 did struggle with certain visual and instrument-based questions. Aashvi shared that it failed to accurately interpret a Vernier scale and took nearly 10 minutes to answer a graphical question, only to get it wrong. 'It was not able to understand the Vernier Scale readings. It kept reiterating to get to the solution but took very long and even then gave the wrong answer,' she wrote.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Mint
a minute ago
- Mint
ChatGPT may be a little different now as OpenAI rolls out a new ‘warmer and friendlier' update
OpenAI has announced that it has made the new GPT-5 AI model more 'warmer and friendlier' following initial feedback that the model felt too formal. The AI startup says that while the changes are subtle, GPT-5–powered ChatGPT should now feel more approachable. Despite adding some light flattery in the model with gestures like 'Good question' and 'Great start,' OpenAI says GPT-5 shows no increase in sycophancy (being overly agreeable with users) compared to the earlier GPT-5 personality. The new update is rolling out to all users and should take about a day to complete. Meanwhile, head of ChatGPT Nick Turley said that users can further customize ChatGPT's personality through the Custom Instructions settings. He also hinted at new upcoming ways to tailor ChatGPT's personality according to user preference. Notably, OpenAI also provides an option to choose from four different personalities — Cynic, Robot, Listener, and Nerd — to tailor GPT-5's responses. However, this feature is currently restricted to paying customers and does not apply to the free tier of the app. OpenAI had high hopes for its GPT-5 rollout, with CEO Sam Altman creating significant hype around the new AI model in the past few months. However, when the model finally launched last week, it was met with anything but a warm reception from users. ChatGPT users complained that GPT-5's responses were shorter and seemed to lack the emotional depth of the previous model. Just a few months ago, OpenAI faced criticism for making GPT-4o too sycophantic, but this time, its new model received backlash for almost the opposite reasons. Another major reason the GPT-5 launch fell flat was user frustration over the removal of all older AI models from ChatGPT. This not only disrupted workflows for many but also reduced usage limits for $20/month ChatGPT Plus subscribers, who went up in arms and threatened to cancel their subscriptions. OpenAI then increased the rate limits for ChatGPT Plus users, first for the standard model and later for the GPT-5 Thinking model as well. The company also brought back the model picker in ChatGPT to allow users to select which AI model should answer their queries.
&w=3840&q=100)
Business Standard
31 minutes ago
- Business Standard
OpenAI staff plan to sell $6 billion in stock to SoftBank, other investors
Current and former OpenAI employees plan to sell approximately $6 billion worth of shares to an investor group that includes Thrive Capital, SoftBank Group Corp. and Dragoneer Investment Group, in a deal that values the ChatGPT maker at $500 billion, according to people familiar with the matter. The talks are early and the size of the share sale could still change, said the people, who asked not to be identified discussing private information. The secondary share investment is on top of SoftBank's commitment to lead OpenAI's $40 billion funding round, which values the company at $300 billion, according to another person familiar with the deal. That round remains ongoing, with OpenAI recently securing $8.3 billion from a syndicate of investors. Representatives for Dragoneer and Thrive didn't respond to requests for comment. Spokespeople for OpenAI and SoftBank declined to comment. All three firms are existing OpenAI backers. Allowing employees to sell shares is an important tool for startups trying to retain top talent, without requiring the company to go public or be acquired. In some cases, early investors also use these deals to sell down their stakes, though OpenAI investors are not eligible to do so in this round, according to a person familiar with the matter. Current and former employees who spent at least two years at the company are able to participate. With its participation in the share sale, as well as its previous commitments, SoftBank is making a pivotal bet on the success of OpenAI. In addition to those deals, the Japanese conglomerate headed by Masayoshi Son recently closed a separate $1 billion purchase of OpenAI employee shares at a $300 billion valuation, according to a person familiar with the matter. Negotiations for that deal started before talks around the $500 billion secondary valuation began, they said. The $500 billion valuation would make OpenAI the world's most valuable startup, surpassing Elon Musk's SpaceX. The company expects revenue to triple this year to $12.7 billion, up from $3.7 billion in 2024, Bloomberg has reported. And the secondary deal talks come on the heels of the release of its highly-anticipated GPT-5 model. This week, OpenAI chief Sam Altman sat down with a group of reporters and shared his vision for the company, including that it wants to spend trillions of dollars on the infrastructure required to run AI services in the 'not very distant future.' 'You should expect a bunch of economists to wring their hands and say, 'This is so crazy, it's so reckless,' and whatever,' Altman said. 'And we'll just be like, 'You know what? Let us do our thing.''
Mint
an hour ago
- Mint
OpenAI could buy Google Chrome, says Sam Altman — but only if this happens
OpenAI CEO Sam Altman has confirmed that the AI startup is interested in buying Google Chrome if Google is forced to sell its browser by the US government. The confirmation from Altman came just days after a Wall Street Journal report revealed that Perplexity AI had reportedly offered $34.5 billion in cash to acquire Chrome from Alphabet. Notably, Google lost an antitrust trial in the United States last year, where the judge ruled that the tech giant had illegally monopolized the online search market. As part of the remedial measures, the Department of Justice has suggested that Chrome be sold to an independent entity to level the playing field. Speaking to reporters, Sam Altman said (as quoted by TechCrunch), 'If Chrome is really going to sell, we should take a look at it.' 'Is it actually going to sell? I assumed it wasn't gonna happen,' the OpenAI CEO added. Meanwhile, there have been reports that OpenAI is building its own AI browser, and Altman confirmed that the recent addition of its CEO of Applications, Fidji Simo, means the company is working on many consumer applications outside of ChatGPT. Altman also stated that OpenAI is interested in exploring a new social media platform, saying there is 'nothing' inspiring to him about the way AI is used on social media today, TechCrunch reported. He said OpenAI is interested in 'whether or not it is possible to build a much cooler kind of social experience with AI.' The OpenAI CEO also confirmed that the company would be backing a neural interface startup called Merge Labs to compete with Elon Musk's Neuralink. 'I think neural interfaces are cool ideas to explore. I would like to be able to think something and have ChatGPT respond to it,' he added. Notably, tensions have flared up between Musk and Altman over alleged Apple bias over App Store. While Musk accused the iPhone maker of being biased towards ChatGPT and not featuring his Grok AI application in any of its rankings, Altman took a dig at the billionaire over X's own policies of amplifying Musks' tweets to all users.
