Semantics podcasts

Stanford linguist and computer scientist Chris Manning discusses what linguistics contributed to machine learning and why pragmatics and dialogue remain open problems for language models. The conversation examines early abstraction of verb categories in small transformers, how distributed representations shape learning, and whether language alone can support meaningful representations. It also considers language in world models, diffusion language models, and representation finetuning (ReFT), which steers frozen models through their hidden states. The final discussion asks where knowledge resides and whether concepts occupy linear subspaces.

Hosted by Ravid Shwartz Ziv, Allen Roush

Psychologist Steven Pinker joins Shankar Vedantam to explore why people communicate indirectly despite learning that honesty is desirable. The conversation examines hedging, softened wording and meanings left unstated, and considers the role these patterns play in human relationships.

Hosted by Shankar Vedantam