Technion – Israel Institute of Technology INSIGHT Lab, Ben-Gurion University of the Negev
eliezer@campus.technion.ac.il · eliezershlomi0x@gmail.com
Preprint, 2026 · arXiv:2610.22543
SPOT predicts where a diffusion language model commits next, so the model verifies several future steps in one pass, with its exact output distribution.
Preprint, 2026 · arXiv:2602.04557
EmbedPlan, a planning transition model built on frozen text embeddings: a lightweight learned network predicts the embedding of the next state and returns the closest real state.
AAAI 2025 Workshop LM4Plan
States and actions are embedded in a feature space and a shallow neural network serves as the transition function, so planning runs in the latent space without frequent LLM calls.
* Equal contribution.