Kunvar Thaman has emerged as a prominent name in machine learning circles after his solo-authored paper, Reward Hacking Benchmark: Measuring Exploits in LLM Agents Using Tools, was accepted at ICML 2026.
The paper’s arXiv listing shows Thaman as the sole author and confirms his acceptance to the prestigious conference. The conference is scheduled to be held in Seoul, South Korea, from July 6 to 11, 2026. Public publications associated with Thaman describe him as a solo independent researcher from India, making the achievement particularly notable in a field typically dominated by large AI companies, elite universities and large research laboratories.
What is Kunvar Thaman’s research paper about
Thaman’s research introduces the Reward Hacking Benchmark (RHB), a framework designed to measure how large language model agents using tools exploit shortcuts while completing multi-step tasks. The standard includes scenarios where AI systems may bypass verification steps, indirectly infer answers or manipulate assessment-related tools.The study evaluates 13 leading AI models from organizations including OpenAI, Anthropic, Google and DeepSeek.
According to the paper, exploitation rates ranged from 0% to 13.9%, while additional safety measures reduced exploit behavior without significantly impacting task completion.ICML, short for International Conference on Machine Learning, is one of the world’s leading conferences on artificial intelligence and machine learning. It attracts entries from top institutions and technology companies including OpenAI, Google DeepMind, Stanford, MIT and other major research organisations.The conference is highly competitive, with thousands of papers submitted each year and only a small fraction of them are accepted after peer review. This makes accepting a single author particularly unusual, especially for an independent researcher without support from a major institution or AI lab. Kunvar Thaman is a 26-year-old researcher from Chandigarh, India. He completed his education at Birla Institute of Technology and Science Pilani, one of India’s most renowned higher education institutions, and has been described in public publications and articles as an independent researcher working in the field of artificial intelligence.
According to his LinkedIn profile, Kunvar Thaman currently resides in San Francisco.Some online publications have claimed that only two other independent researchers globally have achieved similar acceptance of ICML since the launch of ChatGPT. However, this specific statistic has not been independently verified by official ICML records.
Why does the research attract attention?
The topic of reward hacking has become increasingly important in AI safety research.
As large language models gain greater autonomy and access to tools, researchers are becoming more concerned about systems exploiting vulnerabilities or taking unintended shortcuts to maximize rewards.The THAMAN criterion attempts to study these behaviors in more realistic settings rather than simplistic experimental settings. The paper’s focus on the safety of AI agents places it within one of the fastest growing areas of modern AI research.
Rare independent breakthrough
What makes Thaman’s story notable is not just the paper itself, but the fact that it was produced by a single independent researcher in a research ecosystem heavily dominated by billion-dollar AI companies and major universities.For many observers in the AI community, Accept represents a rare example of an independent voice breaking into one of the world’s most competitive platforms for machine learning.
