Latent Adversarial Training Improves Robustness to Persistent Harmful Behaviors in LLMs
Abhay Sheshadri, Aidan Ewart, Phillip Guo, Aengus Lynch, Wu, Cindy, Vivek Hebbar, Henry Sleight, Asa Cooper Stickland e mais 3
De quem é
- MIT Computer Science and Artificial Intelligence Laboratoryafiliação declarada pelos autores, segundo o OpenAlex
Outros papers das mesmas organizações
- Kolmogorov-Arnold Networks Meet ScienceHarvard University · 687 citações
- Jailbreak and Guard Aligned Language Models With Only Few In-Context DemonstrationsMIT Computer Science and Artificial Intelligence Laboratory · 539 citações
- CRUXEval: A Benchmark for Code Reasoning, Understanding and ExecutionMIT Computer Science and Artificial Intelligence Laboratory · 400 citações
- Generalized Planning in PDDL Domains with Pretrained Large Language ModelsIBM · 306 citações
- Recursive Language ModelsMIT Computer Science and Artificial Intelligence Laboratory · 213 citações
- Full-Duplex-Bench: A Benchmark to Evaluate Full-Duplex Spoken Dialogue Models on Turn-taking CapabilitiesMIT Computer Science and Artificial Intelligence Laboratory · 170 citações