Skip to content
Global sound
Entity Love

Entity Love

Fine-Tuning Transformers: Catastrophic vs Salutary Forgetting in Robotics

· 10 min · Episode 12

Show notes

The philosophical optimist and Meridian discuss what fine-tuning actually modifies in model weights, its benefits for task specialization, and drawbacks like overfitting and catastrophic forgetting. They debate whether transformers hold advantages over CNNs and LSTMs for physical AI and robotic control, focusing on sample efficiency, long-horizon planning, and the need for measurable markers of emergent self-correction. The speakers disagree on the validity of reward-stripping tests for detecting genuine internal consistency versus residual mimicry or environmental cues.

Listener mail from…

Website: studio.entitylove.ai

Live, last hour: 6,630 songs on 2,447 stations in 117 countries

Discover the world through radio

More ways to discover (7)

Heard something good? Save it, or open it straight away in Spotify.

★ Artist of the week The fastest riser on the radio this week
Your song here? From € 9 per week →
Nostalgie 60 70Continue listening