Yuetai Li

large language model (4) adversarial learning (2) chain-of-thought reasoning (2) model safety (1) backdoor attack (1) safety evaluation (1) decoding strategy (1) reasoning model (1) reasoning benchmark (1) reasoning trace (1) attack success rate (1) safety training (1) small language model (1) research agent (1) temporal sampling (1) jailbreak evaluation (1) scientific publishing (1) reasoning capability transfer (1) backdoor attack mitigation (1) inference time defense (1)