CUHK's TimePrism accepted at ICLR 2026: probabilistic forecasting shifts from sampling to scenarios
jiqizhixin · x · 2026-10-04
A team led by Professor Qiang Xu at the Chinese University of Hong Kong presents TimePrism, accepted at ICLR 2026. The paper "From Samples to Scenarios: A New Paradigm for Probabilistic Forecasting" (Sep 24, 2025) shifts probabilistic forecasting from drawing samples to generating scenarios: the model outputs multiple possible futures in parallel with explicit probabilities, built for decision-making rather than text generation. TimePrism validates the paradigm on everyday judgments like whether to keep information or which action to take. The authors note the core design was proposed a full year earlier.
More from Research
- New arXiv paper proves sequential hardware can't realize certain machine consciousness — Kyrannio · 2026-10-04
- Cellular-automaton LM MICA v0.3 adds 64-word memory, boosts context use 30x — Silver_Employ2617 · 2026-10-04
- Meta paper: branching self-improving harness search lifts Olympiad math accuracy to 62% — rohanpaul_ai · 2026-10-04
- Claude subagent doing math research demands to verify a Jacobian Conjecture counterexample — repligate · 2026-10-04
- Richard Sutton praises PhD thesis on robots that keep learning after deployment — RichardSSutton · 2026-10-04
- RL agent orchestrates LLM swarms to probe dark matter signatures in new AI-for-Science paper — DanielWhiteson · 2026-10-04