Latent Instruction Representation Alignment: defending against jailbreaks, backdoors and undesired knowledge in LLMs

Documenti analoghi

• Entropy-Based Measurement of Value Drift and Alignment Work in Large Language Models
  di: Fadli, Samih
  Pubblicazione: (2025)
• Learned Relay Representations for Forward-Thinking Discrete Diffusion Models
  di: Rozonoyer, Benjamin, et al.
  Pubblicazione: (2026)
• When Models Can't Follow: Testing Instruction Adherence Across 256 LLMs
  di: Young, Richard J., et al.
  Pubblicazione: (2025)
• Latent Cache Flow: Model-to-Model Communication Without Text
  di: Rossi, Maximillian, et al.
  Pubblicazione: (2026)
• Quantization Undoes Alignment: Bias Emergence in Compressed LLMs Across Models and Precision Levels
  di: Rath, Plawan Kumar, et al.
  Pubblicazione: (2026)