fine-tuning
12 talks
Your Fine-Tuned Model Is Tech Debt: A 50x ROI House of Cards
Gradient-Free Continual Learning
Improving Agents is a Data Mining Problem
Local Models: Trust, Control, Optimization
Data and Environment Curation for Post-Training LLMs
Ending AI Slop
Persona Engineering: A Field Guide to AI Synthetic Personas
Why Large? Tiny LMs and Agents on Edge and Robotics
Modern Post-Training: A Deep Dive
Your LLM Deception Monitor Is Broken. The Fix Is in the Training Data
The Miranda Hypothesis: How Hamilton Poisoned Persona Evals
Stop Making Models Bigger, Make Them Behave