Souradip Chakraborty

reinforcement learning from human feedback (2) language model (2) inference-time alignment (2) model-based reinforcement learning (2) safety alignment (2) ensemble learning (2) jailbreak attack (2) adversarial attack (2) text mining (1) policy gradient (1) data poisoning (1) information retrieval (1) text generation (1) posterior estimation (1) text classification (1) backdoor attack (1) bayesian regret (1) question answering (1) optimal decoding (1) reward function (1)