����������

The Distinguished Speaker Series invites distinguished scholars to present their work in the history and philosophy of science. The Allan Franklin New Ideas Speaker Series invites scholars to present their new and exciting work in the history and philosophy of science. It is named in honor of Allan Franklin, for his decades of unparalleled contributions to the history and philosophy of science. Dr. Franklin is best known for his work on the methods of physics.

Upcoming Schedule for AY 25/26:

• Friday, February 27th, 3:30pm, HLMS E210: New Ideas Speaker, (Northeastern), "Against the Platonic Representation Hypothesis"
  • Abstract: The Platonic Representation Hypothesis posits that “neural networks, trained with different objectives on different data and modalities, are converging to a shared statistical model of reality in their representation spaces” (Huh et al. 2024).  This realist explanation for the success of contemporary large language and image models has gained traction within computer science, particularly as an explanation for some of the striking large model convergence results found using mechanistic interpretability techniques.  In this talk, I undermine the Platonic Representation Hypothesis by positing alternative explanations for the phenomenon of model convergence and propose a different test of meaningful convergence to replace it.  
• Friday, March 6th, 3:30pm, HLMS E210: Distinguished Speaker, (UC, San Diego), "An Altitude or Attitude Adjustment? Gravity's Challenge to Thermodynamics"
  • Abstract: Consider a simple column of gas extended vertically in a gravitational field on Earth. Make the column tall enough so that gravity matters. Is the gas at the top colder than the gas at the bottom (in equilibrium)? Surprisingly, this question was the subject of debate by Maxwell, Guthrie, Burbury, Boltzmann, and Loschmidt, illustrating that simple systems can raise deep foundational puzzles. The debate was resolved in favor of no temperature gradients arising. In an exciting but under-explored twist to this history, general relativity predicts the opposite. How do we reconcile this Tolman-Ehrenfest effect with the historical resolution? Do we need an altitude adjustment to temperature, as is commonly thought, or do we instead require an attitude adjustment to our understanding of thermodynamics?