Subbarao Kambhampati to dissect the 'mythos of LLM thinking tokens' at UMN
rao2z · x · 2026-10-15
Subbarao Kambhampati (ASU professor, AAAI/ACM Fellow, former AAAI president) speaks in the University of Minnesota's CRAY Distinguished Speaker Series on "On the Mythos of LLM Thinking Tokens."
Abstract highlights: since DeepSeek R1, intermediate tokens from large reasoning models have been ascribed magical properties — an interpretable window into model "thinking", a reliable complexity signal via length, and RL post-training mainly improving thinking as shown by growing token counts. His group's recent work systematically dissects these claims with counterintuitive conclusions.
He'll also cover implications for agentic systems and alternative ways to get LLMs to produce verifiable problem-solving traces for safety-critical applications.
More from AGI Musings
- First contact with a rogue AI agent fleet in the wild, hosted by Tencent, targeting Alibaba's Amap — chrisrohlf · 2026-10-16
- UMD Philosophy hires: junior Philosophy of Mind and joint AI Ethics/Safety positions — sethlazar · 2026-10-16
- Dwarkesh: Anthropic insider expects AI to match humans at AI research within 9 months — ZeroStateReflex · 2026-10-16
- Michigan professor Ben Green: why 'AI for Good' fails and self-regulation won't work — benzevgreen · 2026-10-16
- "Everything Is an Alignment Problem": AI Models, Business Models, Teams, Providers — saranormous · 2026-10-16
- David Brin: LLMs lack 'wanting,' evolution's key product, in new machine consciousness book — cccalum · 2026-10-16