Indranil Halder's Home page
Indranil Halder is a Research Associate at the Harvard John A. Paulson School of Engineering and Applied Sciences. Previously, Indranil was the Harvard Quantum Initiative Fellow at the Center for the Fundamental Laws of Nature. He is a machine learning researcher specializing in generative AI, with core interests in the interpretability of knowledge representations and the safety and robustness of deep learning systems.
Research
Machine learning
Large language models are now deployed in a wide range of high-stakes applications, which makes a precise understanding of their failure modes as important as continued improvements in their capabilities. Indranil’s research in machine learning aims to identify and quantify the mechanisms by which these systems can fail or be exploited. His approach is to develop analytically tractable theoretical models that yield quantitative predictions, which are then tested in experiments with large language models.
Data poisoning. Web-scale pre-training corpora cannot be curated exhaustively, and a small fraction of corrupted data may therefore be incorporated into training. Existing studies have focused largely on targeted backdoors and their persistence through safety post-training. Indranil and collaborators address a more fundamental question: how performance on clean data degrades as a function of the poison rate. Controlled pre-training experiments exhibit power-law degradation with a non-integer exponent, and the work develops a solvable theory that accounts for this behavior.
Reward hacking. The performance of language models is increasingly improved at inference time by generating multiple candidate responses and selecting among them with a judge model. Because the judge is an imperfect proxy for response quality, additional search may exploit its deficiencies rather than improve the underlying output, a phenomenon known as reward hacking. Indranil’s analytically tractable model of the LLM-as-a-Judge setting characterizes the conditions under which additional inference-time computation yields monotonic improvement and those under which there exists a finite optimal number of samples.
Jailbreaking attacks. Repeated sampling also amplifies safety failures, since a small per-generation probability of an unsafe response accumulates over many attempts. Indranil and collaborators demonstrate that the probability of a successful jailbreak obeys qualitatively distinct scaling laws, polynomial or exponential in the number of samples, depending on the strength of the attack. The crossover between these regimes is explained by a random graph model of binary language generation.
Long context failure. The reliability of deployed systems further depends on their ability to retrieve and use relevant information from increasingly long contexts. Language models exhibit a systematic positional bias and frequently fail to use information located in the middle of the input, a behavior known as the “lost in the middle” phenomenon. A theoretical understanding of this bias is essential for systems that must recall critical facts and maintain persistent memory without becoming susceptible to hallucination. Indranil is also interested in, and actively working on, related topics.
Theoretical physics
Before transitioning to machine learning, Indranil made notable contributions to theoretical physics. During his early days at Harvard University, Indranil established a new strong-weak duality closely related to ER=EPR - the fascinating connection between entanglement and geometry. Using the duality, he has made distinct progress on the long-standing issue of thermal microstate counting of blackholes within string theory. He also played a leading role in the study of supersymmetric blackholes and blackrings in M-theoretic, F-theoretic compactification. In addition, he proposed a framework to evaluate the supersymmetric index in disordered quantum field theories in terms of bi-local fields closely related to wormhole-like physics.