Compositional Literary Primitives in Instruction-Tuned LLMs: Cross-Architectural SAE Features for Self, Style, and Affect

Documenti analoghi

• Call2Instruct: Automated Pipeline for Generating Q&A Datasets from Call Center Recordings for LLM Fine-Tuning
  di: Echeverria, Alex, et al.
  Pubblicazione: (2025)
• How Good Are LLMs for Literary Translation, Really? Literary Translation Evaluation with Humans and LLMs
  di: Zhang, Ran, et al.
  Pubblicazione: (2024)
• Dense SAE Latents Are Features, Not Bugs
  di: Sun, Xiaoqing, et al.
  Pubblicazione: (2025)
• CrossIn: An Efficient Instruction Tuning Approach for Cross-Lingual Knowledge Alignment
  di: Lin, Geyu, et al.
  Pubblicazione: (2024)
• SelectIT: Selective Instruction Tuning for LLMs via Uncertainty-Aware Self-Reflection
  di: Liu, Liangxin, et al.
  Pubblicazione: (2024)