LLMs produce inconsistent probabilities: P(rain) + P(no rain) don't sum to one
soumitrashukla9 · x · 2026-09-05
Suproteem Sarkar gives a concrete illustration of LLM calibration problems: ask a model the probability of rain tomorrow and you may get P(rain)=0.7, then ask the probability it won't rain and get 0.2 — the two never sum to one, violating basic probability axioms. The work is praised as some of the most interesting AI+economics research, examining consistency of model-generated probabilities.
Related event: LLMs give incoherent probabilities: P(rain) + P(no rain) ≠ 1(2 posts)→
More from Research
- TAOCP open problems released as a dataset to benchmark frontier models — sytelus · 2026-09-05
- Clinic-in-the-Loop: why clinical trials are the real bottleneck breaking Eroom's Law — anshulkundaje · 2026-09-05
- Contentious preprint claims ensemble-wrapped LLM achieves phenomenal consciousness — PeterBowdenLive · 2026-09-05
- Fine-tuning on 'random' numbers transfers teacher bias — subliminal learning worries — ryanorban · 2026-09-05
- Survey of 668 developers: readers who suspect AI writing will stop reading and block you — IanArawjo · 2026-09-05
- Randomized YaRN: training on short context with sampled positions boosts 128K reasoning — gregd_nlp · 2026-09-05